SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2304.08485 · NeurIPS · 2023

Visual Instruction Tuning

Chunyuan Li, Haotian Liu, Qingyang Wu, Yong Jae Lee

arXiv · PDF · Open in the Atlas

Code that ran

16 of 51 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (6 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 7 have no contract verdict; 8 failures of our instrument; 35 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2304.08485.json · badge: /api/ran/2304.08485.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
haotian-liu/LLaVA canonical (classifier label)
classifier basis: classifier_2025
4 ran · 8 no recorded run
llava-annonymous/llava canonical (classifier label)
classifier basis: classifier_2025
2 ran · 3 no recorded run
LLaVA-VL/LLaVA-NeXT canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
dinhvietcuong1996/icme25-inova community
Papers with Code archive
5 ran · 7 no recorded run
sunsmarterjie/chatterbox community
Papers with Code archive
1 ran · 2 no recorded run
sshh12/multi_token community
Papers with Code archive
1 ran · 0 no recorded run
ictnlp/LLaVA-Mini extension (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
ZhangYiqun018/StickerConv community
Papers with Code archive
0 ran · 5 no recorded run
skunkworksai/bakllava community
Papers with Code archive
0 ran · 3 no recorded run
tabtoyou/kollava community
Papers with Code archive
0 ran · 2 no recorded run
LLaVA-VL/LLaVA-Plus-Codebase extension (classifier label)
classifier basis: classifier_2025
0 ran · 2 no recorded run
PKU-YuanGroup/MoE-LLaVA extension (classifier label)
classifier basis: classifier_2025
0 ran · 2 no recorded run
FunctionStatusWhere it lives
IdentityMap Ran
ran on a synthesized input
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("1707c6987676247e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

INIT_ARGS = {}

ARGS = {
    'x': torch.randn(2, 4)
}

# Construct and call
model = torch.nn.Module.__new__(type('IdentityMap', (torch.nn.Module,), {
    '__init__': lambda self: torch.nn.Module.__init__(self),
    'forward': lambda self, x, *args, **kwargs: x,
    'config': property(lambda self: {'mm_resampler_type': None})
}))
model.__init__()

output = model(ARGS['x'])
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
LlavaConfig Ran
ran on a synthesized input · constructed an object; no result was computed
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("dbe52eb96f05b6db")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, LlavaConfig: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from transformers import LlamaConfig

torch.manual_seed(0)

# LlavaConfig inherits from LlamaConfig, so we need to provide all LlamaConfig parameters
# plus the LlavaConfig-specific ones

ARGS = {
    # LlamaConfig parameters (required)
    "vocab_size": 32000,
    "hidden_size": 4096,
    "intermediate_size": 11008,
    "num_hidden_layers": 2,
    "num_attention_heads": 32,
    "num_key_value_heads": 8,
    "hidden_act": "silu",
    "max_position_embeddings": 2048,
    "initializer_range": 0.02,
    "rms_norm_eps": 1e-6,
    "use_cache": True,
    "pad_token_id": 0,
    "bos_token_id": 1,
    "eos_token_id": 2,
    "pretraining_tp": 1,
    "tie_word_embeddings": False,
    "rope_theta": 10000.0,
    "rope_scaling": None,
    "attention_bias": False,
    "attention_dropout": 0.0,
    
    # LlavaConfig-specific parameters
    "temperature": 0.7,
    "max_new_tokens": 512,
    "do_sample": True,
    "top_p": 0.9,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type LlavaConfig — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a LlavaConfig output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
LlavaConfig Ran
ran on a synthesized input · constructed an object; no result was computed
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("f01242335a1498b3")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, LlavaConfig: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from transformers import LlamaConfig

torch.manual_seed(0)

# LlamaConfig inherits from PretrainedConfig which has many parameters
# We need to provide values for all parameters that LlamaConfig.__init__ accepts

ARGS = {
    "vocab_size": 32000,
    "hidden_size": 768,
    "intermediate_size": 3072,
    "num_hidden_layers": 4,
    "num_attention_heads": 8,
    "num_key_value_heads": 8,
    "hidden_act": "silu",
    "initializer_range": 0.02,
    "initializer_factor": 1.0,
    "rms_norm_eps": 1e-6,
    "use_cache": True,
    "pad_token_id": 0,
    "bos_token_id": 1,
    "eos_token_id": 2,
    "pretraining_tp": 1,
    "tie_word_embeddings": False,
    "rope_scaling": None,
    "attention_dropout": 0.0,
    "max_position_embeddings": 2048,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type LlavaConfig — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a LlavaConfig output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
LlavaConfig Ran
ran on a synthesized input · constructed an object; no result was computed
llava-annonymous/llava/llava/model/llava.py
code served (permissive licence) · get_code("92cb18f9e27a4e39")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, LlavaConfig: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from transformers import LlamaConfig

torch.manual_seed(0)

# LlavaConfig inherits from LlamaConfig, so we need to provide all LlamaConfig parameters
# Based on transformers library, LlamaConfig accepts these core parameters:

ARGS = {
    # LlamaConfig / PretrainedConfig parameters
    "vocab_size": 32000,
    "hidden_size": 4,
    "intermediate_size": 8,
    "num_hidden_layers": 2,
    "num_attention_heads": 2,
    "num_key_value_heads": 2,
    "hidden_act": "silu",
    "initializer_range": 0.02,
    "initializer_factor": 1.0,
    "rms_norm_eps": 1e-6,
    "use_cache": True,
    "pad_token_id": 0,
    "bos_token_id": 1,
    "eos_token_id": 2,
    "pretraining_tp": 1,
    "tie_word_embeddings": False,
    "rope_theta": 10000.0,
    "rope_scaling": None,
    "attention_bias": False,
    "attention_dropout": 0.0,
    # Base PretrainedConfig parameters
    "architectures": ["LlavaForCausalLM"],
    "model_type": "llava",
    "task_specific_params": None,
    "problem_type": None,
    "id2label": None,
    "label2id": None,
    "max_position_embeddings": 2048,
    "output_past": False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type LlavaConfig — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a LlavaConfig output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
LlavaLlamaModel Ran
ran on a synthesized input · constructed an object; no result was computed
sunsmarterjie/chatterbox/model/llava/model/llava.py
code served (permissive licence) · get_code("e6460881f1cf311c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, LlavaLlamaModel: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from transformers import LlamaConfig

torch.manual_seed(0)

# Create a minimal LlamaConfig
config = LlamaConfig(
    vocab_size=100,
    hidden_size=128,
    intermediate_size=256,
    num_hidden_layers=2,
    num_attention_heads=2,
    max_position_embeddings=512,
)

# Create mock vision tower and related components
class MockVisionConfig:
    def __init__(self):
        self.image_size = 224
        self.patch_size = 16
        self.hidden_size = 256

class MockVisionModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.config = MockVisionConfig()
        self.linear = nn.Linear(256, 256)
    
    def forward(self, x, output_hidden_states=False):
        batch_size = x.shape[0]
        # Return mock hidden states with proper structure
        # (batch_size, num_patches + 1, hidden_size)
        num_patches = (self.config.image_size // self.config.patch_size) ** 2
        hidden_state = torch.randn(batch_size, num_patches + 1, self.config.hidden_size)
        
        class Output:
            def __init__(self, hs):
                self.hidden_states = [hs, hs]
        
        return Output(hidden_state)
    
    def requires_grad_(self, requires_grad):
        for param in self.parameters():
            param.requires_grad = requires_grad
        return self
    
    def to(self, dtype):
        return self

# Construct ARGS dictionary
ARGS = {
    'config': config,
    'mm_vision_tower': None,
    'mm_hidden_size': None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type LlavaLlamaModel — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a LlavaLlamaModel output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
SimpleResBlock Ran
ran on a synthesized input · constructed an object; no result was computed
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("4a7e5ec29fb79acb")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, SimpleResBlock: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "channels": 4
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SimpleResBlock — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a SimpleResBlock output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
_MLPVectorProjector Ran
ran on a synthesized input · constructed an object; no result was computed
sshh12/multi_token/multi_token/modalities/projectors.py
code served (permissive licence) · get_code("078a686c14a5d016")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, _MLPVectorProjector: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "input_hidden_size": 4,
    "lm_hidden_size": 5,
    "num_layers": 3,
    "width": 2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type _MLPVectorProjector — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a _MLPVectorProjector output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
build_vision_projector Ran
ran; our contract draft was wrong, not the code
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("ea7ba84c548c5a33")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Linear. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
import re

torch.manual_seed(0)

class SimpleResBlock(nn.Module):
    def __init__(self, channels):
        super().__init__()
        self.pre_norm = nn.LayerNorm(channels)
        self.proj = nn.Sequential(nn.Linear(channels, channels), nn.GELU(), nn.Linear(channels, channels))

    def forward(self, x):
        x = self.pre_norm(x)
        return x + self.proj(x)

class IdentityMap(torch.nn.Module):
    def __init__(self):
        super().__init__()

    def forward(self, x, *args, **kwargs):
        return x

    @property
    def config(self):
        return {'mm_resampler_type': None}

class MockVisionConfig:
    def __init__(self):
        self.hidden_size = 8
        self.image_size = 224

class MockConfig:
    def __init__(self, projector_type='linear'):
        self.mm_projector_type = projector_type
        self.mm_hidden_size = 4
        self.hidden_size = 8

ARGS = {
    'config': MockConfig('linear'),
    'delay_load': False,
    'vision_cfg': MockVisionConfig()
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Linear — a Linear output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Linear output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
build_vision_resampler Ran
ran; our contract draft was wrong, not the code
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("10f2142755f4bbb5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned IdentityMap. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class MockModelArgs:
    """Stand-in for model_args that has mm_resampler_type attribute."""
    def __init__(self, resampler_type=None):
        self.mm_resampler_type = resampler_type

ARGS = {
    'model_args': MockModelArgs(resampler_type=None),
    'delay_load': False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type IdentityMap — a IdentityMap output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a IdentityMap output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
convert_caps Ran
ran; our contract draft was wrong, not the code
llava-annonymous/llava/llava/eval/eval_science_qa.py
code served (permissive licence) · get_code("94fda0761ce0d5b9")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Construct valid arguments for convert_caps function
results = [
    {
        'question_id': 42,
        'text': 'A dog running in the park'
    },
    {
        'question_id': 123,
        'text': 'A cat sitting on a mat'
    }
]

ARGS = {
    'results': results
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_anyres_image_grid_shape Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("2ff0d742a84cb8e2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import ast

# Construct valid arguments for get_anyres_image_grid_shape

# image_size: tuple of (width, height)
image_size = (512, 384)

# grid_pinpoints: can be either a list or a string representation of a list
# Using a list of possible resolutions (width, height tuples)
grid_pinpoints = [(256, 192), (512, 384), (1024, 768)]

# patch_size: integer representing the size of each image patch
patch_size = 16

ARGS = {
    "image_size": image_size,
    "grid_pinpoints": grid_pinpoints,
    "patch_size": patch_size,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family int — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
image_parser Ran
ran; our contract draft was wrong, not the code
ictnlp/LLaVA-Mini/llavamini/eval/run_llava_mini.py
code served (permissive licence) · get_code("a7bee88c1c7fd6a3")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

class Args:
    def __init__(self):
        self.image_file = "image1.jpg,image2.png,image3.bmp"
        self.sep = ","

ARGS = {"args": Args()}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
should_only_save_mm_adapter Ran
ran, violated the contract we drafted
LLaVA-VL/LLaVA-NeXT/llava/train/llava_trainer.py
code served (permissive licence) · get_code("08787a49ef19e0aa")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned bool, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class Args:
    def __init__(self):
        self.lora_enable = False
        self.tune_mm_mlp_adapter = False
        self.mm_tunable_parts = "mm_mlp_adapter"

ARGS = {
    'args': Args()
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
split_to_even_chunks Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
haotian-liu/LLaVA/llava/train/llava_trainer.py
code served (permissive licence) · get_code("10893c4608c08075")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned list, ndim 2, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Analyzing the function:
# - indices: a list/sequence of indices that will be iterated over
# - lengths: a dict-like or list-like object that maps indices to their lengths (accessed via lengths[index])
# - num_chunks: an integer specifying how many chunks to split into

# Create sample data
indices = [0, 1, 2, 3, 4, 5]
lengths = {0: 10, 1: 15, 2: 8, 3: 12, 4: 9, 5: 11}
num_chunks = 3

ARGS = {
    "indices": indices,
    "lengths": lengths,
    "num_chunks": num_chunks,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
unpad_image Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
haotian-liu/LLaVA/llava/model/llava_arch.py
code served (permissive licence) · get_code("55c32993da87759b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 3, dtype family float, shape (3, 10, 12) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a padded image tensor in CxHxW format
# Let's say original image was 8x6 (width x height)
# After padding and resizing, it became 12x12 (square)
# C=3 (RGB), H=12, W=12
tensor = torch.randn(3, 12, 12)

# Original size as (width, height) tuple
original_size = (8, 6)

ARGS = {
    "tensor": tensor,
    "original_size": original_size,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (3, 10, 12) — shape (3, 10, 12) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
unpad_image Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
LLaVA-VL/LLaVA-NeXT/llava/model/llava_arch.py
code served (permissive licence) · get_code("7606525af238fb64")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 3, dtype family float, shape (3, 8, 6) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a padded image tensor in CxHxW format
# C=3 (RGB), H=8 (current height), W=6 (current width)
tensor = torch.randn(3, 8, 6)

# Original size as (width, height)
# Using different dimensions to be distinct: width=4, height=5
original_size = (4, 5)

ARGS = {
    "tensor": tensor,
    "original_size": original_size,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (3, 8, 6) — shape (3, 8, 6) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
CLIPVisionTower Not yet run
unverified (harvested, no recorded run)
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("98e6fcd805f498d1")
How it ran

no invocation recorded.

CLIPVisionTowerS2 Not yet run
unverified (harvested, no recorded run)
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("d7be74cc1bd52de7")
How it ran

no invocation recorded.

CLIPVisionTowerS2 Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("5d1c9ab7042cee8d")
How it ran

no invocation recorded.

LlavaConfig Not yet run
unverified (harvested, no recorded run)
sunsmarterjie/chatterbox/model/llava/model/llava.py
code served (permissive licence) · get_code("d2904a880e39bfc2")
How it ran

no invocation recorded.

LlavaLlamaForCausalLM Not yet run
unverified (harvested, no recorded run)
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("1717d5d0a430f17f")
How it ran

no invocation recorded.

LlavaLlamaForCausalLM Not yet run
unverified (harvested, no recorded run)
ZhangYiqun018/StickerConv/sticker_process/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("44969b8364ff17f1")
How it ran

no invocation recorded.

LlavaLlamaForCausalLM Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("4363aef6226bb213")
How it ran

no invocation recorded.

LlavaLlamaForCausalLM Not yet run
unverified (harvested, no recorded run)
llava-annonymous/llava/llava/model/llava.py
code served (permissive licence) · get_code("7adeedb5a7dcbd7a")
How it ran

no invocation recorded.

LlavaLlamaForCausalLM Not yet run
unverified (harvested, no recorded run)
skunkworksai/bakllava/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("e589073e81d58266")
How it ran

no invocation recorded.

LlavaLlamaForCausalLM Not yet run
unverified (harvested, no recorded run)
sunsmarterjie/chatterbox/model/llava/model/llava.py
code served (permissive licence) · get_code("d138a906d3799efb")
How it ran

no invocation recorded.

LlavaLlamaForCausalLM Not yet run
unverified (harvested, no recorded run)
tabtoyou/kollava/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("d70c330590b6be51")
How it ran

no invocation recorded.

LlavaLlamaModel Not yet run
unverified (harvested, no recorded run)
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("a39a4ca102cca6cc")
How it ran

no invocation recorded.

LlavaLlamaModel Not yet run
unverified (harvested, no recorded run)
ZhangYiqun018/StickerConv/sticker_process/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("7fc60199325c39f5")
How it ran

no invocation recorded.

LlavaLlamaModel Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("9707b7b56a48db85")
How it ran

no invocation recorded.

LlavaLlamaModel Not yet run
unverified (harvested, no recorded run)
llava-annonymous/llava/llava/model/llava.py
code served (permissive licence) · get_code("e992c05c75e7c542")
How it ran

no invocation recorded.

LlavaLlamaModel Not yet run
unverified (harvested, no recorded run)
skunkworksai/bakllava/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("d9cce1fb106ab082")
How it ran

no invocation recorded.

LlavaMetaForCausalLM Not yet run
unverified (harvested, no recorded run)
LLaVA-VL/LLaVA-NeXT/llava/model/llava_arch.py
code served (permissive licence) · get_code("39343c26374462af")
How it ran

no invocation recorded.

LlavaMetaForCausalLM Not yet run
unverified (harvested, no recorded run)
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("31bedbeb8856618e")
How it ran

no invocation recorded.

LlavaMetaForCausalLM Not yet run
unverified (harvested, no recorded run)
ZhangYiqun018/StickerConv/sticker_process/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("abf7b5851a13bc76")
How it ran

no invocation recorded.

LlavaMetaForCausalLM Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("8e3647258cae7507")
How it ran

no invocation recorded.

LlavaMetaForCausalLM Not yet run
unverified (harvested, no recorded run)
tabtoyou/kollava/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("2efb2d959762f4ed")
How it ran

no invocation recorded.

LlavaMetaModel Not yet run
unverified (harvested, no recorded run)
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("7732a5ae006ababd")
How it ran

no invocation recorded.

LlavaMetaModel Not yet run
unverified (harvested, no recorded run)
ZhangYiqun018/StickerConv/sticker_process/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("38d280bea74c1e01")
How it ran

no invocation recorded.

LlavaMetaModel Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("0bfdaed6a43a5263")
How it ran

no invocation recorded.

LlavaMetaModel Not yet run
unverified (harvested, no recorded run)
skunkworksai/bakllava/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("5990c8af59a1551f")
How it ran

no invocation recorded.

build_vision_tower Not yet run
unverified (harvested, no recorded run)
dinhvietcuong1996/icme25-inova/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("875ebbe8f3a0c711")
How it ran

no invocation recorded.

build_vision_tower Not yet run
unverified (harvested, no recorded run)
ZhangYiqun018/StickerConv/sticker_process/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("014050859277825c")
How it ran

no invocation recorded.

build_vision_tower Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/model/language_model/llava_llama.py
code served (permissive licence) · get_code("3ee345b55cc88a11")
How it ran

no invocation recorded.

convert_caps Not yet run
unverified (harvested, no recorded run)
LLaVA-VL/LLaVA-Plus-Codebase/llava/eval/eval_science_qa.py
code served (permissive licence) · get_code("150a20daa039f495")
How it ran

no invocation recorded. this sample carries JSON test cases instead of a setup block; get_reference_implementation serves them with the level they back.

get_mm_adapter_state_maybe_zero_3 Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/train/llava_trainer.py
code served (permissive licence) · get_code("bb35e3ac741bb2c9")
How it ran

no invocation recorded.

get_pred_idx Not yet run
unverified (harvested, no recorded run)
LLaVA-VL/LLaVA-Plus-Codebase/llava/eval/eval_science_qa.py
code served (permissive licence) · get_code("63959934f6d1fa06")
How it ran

no invocation recorded. this sample carries JSON test cases instead of a setup block; get_reference_implementation serves them with the level they back.

get_pred_idx Not yet run
unverified (harvested, no recorded run)
llava-annonymous/llava/llava/eval/eval_science_qa.py
code served (permissive licence) · get_code("268b6194c5c23103")
How it ran

no invocation recorded. this sample carries JSON test cases instead of a setup block; get_reference_implementation serves them with the level they back.

maybe_zero_3 Not yet run
unverified (harvested, no recorded run)
haotian-liu/LLaVA/llava/train/llava_trainer.py
code served (permissive licence) · get_code("735025744c1ab0cf")
How it ran

no invocation recorded.

save_image_to_local Not yet run
unverified (harvested, no recorded run)
PKU-YuanGroup/MoE-LLaVA/moellava/serve/gradio_web_server.py
code served (permissive licence) · get_code("e2e5c3a95a6aebb4")
How it ran

no invocation recorded.

save_video_to_local Not yet run
unverified (harvested, no recorded run)
PKU-YuanGroup/MoE-LLaVA/moellava/serve/gradio_web_server.py
code served (permissive licence) · get_code("22930d58d32e2d08")
How it ran

no invocation recorded.

Also present in other papers' code

Of 51 samples lifted from this paper's linked repositories, 51 carry a function identity and 0 do not. Of the 43 distinct functions with an identity, 29 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 3 the same month, 29 later) and 14 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
unpad_image
brick of this paper's assembly: helper
69 other papers
69 later · 5 only as an assembly brick
2309.09958 (2023-09)
An Empirical Study of Scaling Instruct-Tuned Large Multimodal Models
2310.03744 (2023-10)
Improved Baselines with Visual Instruction Tuning
2312.17243 (2023-12)
Unsupervised Universal Image Segmentation
and 66 more not named here
get_mm_adapter_state_maybe_zero_3 43 other papers
1 the same month, 42 later
2304.05884 (2023-04)
Unicom: Universal and Compact Representation Learning for Image Retrieval
2310.03744 (2023-10)
Improved Baselines with Visual Instruction Tuning
2311.04257 (2023-11)
mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaboration
and 40 more not named here
maybe_zero_3 42 other papers
1 the same month, 41 later
2304.05884 (2023-04)
Unicom: Universal and Compact Representation Learning for Image Retrieval
2310.03744 (2023-10)
Improved Baselines with Visual Instruction Tuning
2311.04257 (2023-11)
mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaboration
and 39 more not named here
split_to_even_chunks 40 other papers
1 the same month, 39 later
2304.05884 (2023-04)
Unicom: Universal and Compact Representation Learning for Image Retrieval
2310.03744 (2023-10)
Improved Baselines with Visual Instruction Tuning
2311.04257 (2023-11)
mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaboration
and 37 more not named here
image_parser 16 other papers
16 later
2311.05437 (2023-11)
LLaVA-Plus: Learning to Use Tools for Creating Multimodal Agents
2024.findings-emnlp.268 (2024 (year only))
Math-LLaVA: Bootstrapping Mathematical Reasoning for Multimodal Large Language Models
2404.10618 (2024-04)
Private Attribute Inference from Images with Vision-Language Models
and 13 more not named here
LlavaLlamaModel
brick of this paper's assembly: helper, model
13 other papers
13 later · 3 only as an assembly brick
2404.05673 (2024-04)
CoReS: Orchestrating the Dance of Reasoning and Segmentation
2405.14974 (2024-05)
LOVA 3 : Learning to Visual Question Answering, Asking and Assessment
2406.10701 (2024-06)
MIND: Multimodal Shopping Intention Distillation from Large Vision-language Models for E-commerce Purchase Understanding
and 10 more not named here
LlavaConfig
brick of this paper's assembly: helper
12 other papers
12 later · 9 only as an assembly brick
2309.17102 (2023-09)
Guiding Instruction-based Image Editing via Multimodal Large Language Models
2402.18695 (2024-02)
Grounding Language Models for Visual Entity Recognition
2403.04652 (2024-03)
Yi: Open Foundation Models by 01.AI
and 9 more not named here
build_vision_tower
brick of this paper's assembly: helper
9 other papers
1 dated earlier, 8 later · 4 only as an assembly brick
2301.12597 (2023-01)
BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
2402.11684 (2024-02)
ALLaVA: Harnessing GPT4V-Synthesized Data for Lite Vision-Language Models
2402.18695 (2024-02)
Grounding Language Models for Visual Entity Recognition
and 6 more not named here
save_image_to_local 7 other papers
7 later
Jin_Chat-UniVi_Unified_Visual_Representation_Empowers_Large_Language_Models_with_Image_CVPR_2024_paper (2024 (year only))
Chat-UniVi: Unified Visual Representation Empowers Large Language Models with Image and Video Understanding
2401.06071 (2024-01)
GroundingGPT: Language Enhanced Multi-modal Grounding Model
2401.15947 (2024-01)
MoE-LLaVA: Mixture of Experts for Large Vision-Language Models
and 4 more not named here
save_video_to_local 7 other papers
7 later
Jin_Chat-UniVi_Unified_Visual_Representation_Empowers_Large_Language_Models_with_Image_CVPR_2024_paper (2024 (year only))
Chat-UniVi: Unified Visual Representation Empowers Large Language Models with Image and Video Understanding
2401.06071 (2024-01)
GroundingGPT: Language Enhanced Multi-modal Grounding Model
2401.15947 (2024-01)
MoE-LLaVA: Mixture of Experts for Large Vision-Language Models
and 4 more not named here
get_anyres_image_grid_shape
brick of this paper's assembly: helper
5 other papers
5 later · 4 only as an assembly brick
2406.02539 (2024-06)
Parrot: Multilingual Visual Instruction Tuning
2406.20095 (2024-06)
LLaRA: Supercharging Robot Learning Data for Vision-Language Policy
2408.08849 (2024-08)
ECG-Chat: A Large ECG-Language Model for Cardiac Disease Diagnosis
and 2 more not named here
build_vision_tower
brick of this paper's assembly: helper
4 other papers
4 later · 2 only as an assembly brick
2406.20095 (2024-06)
LLaRA: Supercharging Robot Learning Data for Vision-Language Policy
2502.13508 (2025-02)
VLAS: Vision-Language-Action Model With Speech Instructions For Customized Robot Manipulation
2503.04006 (2025-03)
DSV-LFS: Unifying LLM-Driven Semantic Cues with Visual Features for Robust Few-Shot Segmentation
and 1 more not named here
CLIPVisionTowerS2
brick of this paper's assembly: helper
3 other papers
3 later · 2 only as an assembly brick
2406.20095 (2024-06)
LLaRA: Supercharging Robot Learning Data for Vision-Language Policy
2502.13508 (2025-02)
VLAS: Vision-Language-Action Model With Speech Instructions For Customized Robot Manipulation
2603.12930 (2026-03)
Rethinking VLMs for Image Forgery Detection and Localization
LlavaLlamaForCausalLM
brick of this paper's assembly: entry
3 other papers
3 later
2405.14974 (2024-05)
LOVA 3 : Learning to Visual Question Answering, Asking and Assessment
2410.06234 (2024-10)
TEOChat: A Large Vision-Language Assistant for Temporal Earth Observation Data
2505.00275 (2025-05)
AdCare-VLM: Towards a Unified and Pre-aligned Latent Representation for Healthcare Video Understanding
convert_caps 3 other papers
3 later
2403.09792 (2024-03)
Images are Achilles' Heel of Alignment: Exploiting Visual Vulnerabilities for Jailbreaking Multimodal Large Language Models
2412.00876 (2024-12)
Dynamic-LLaVA: Efficient Multimodal Large Language Models via Dynamic Vision-language Context Sparsification
2503.15621 (2025-03)
LLaVA-MORE: A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction Tuning
get_pred_idx 3 other papers
3 later
2403.09792 (2024-03)
Images are Achilles' Heel of Alignment: Exploiting Visual Vulnerabilities for Jailbreaking Multimodal Large Language Models
2412.00876 (2024-12)
Dynamic-LLaVA: Efficient Multimodal Large Language Models via Dynamic Vision-language Context Sparsification
2503.15621 (2025-03)
LLaVA-MORE: A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction Tuning
LlavaConfig
brick of this paper's assembly: helper
2 other papers
2 later · 2 only as an assembly brick
2406.20095 (2024-06)
LLaRA: Supercharging Robot Learning Data for Vision-Language Policy
2503.04006 (2025-03)
DSV-LFS: Unifying LLM-Driven Semantic Cues with Visual Features for Robust Few-Shot Segmentation
LlavaMetaForCausalLM
brick of this paper's assembly: model
2 other papers
2 later · 2 only as an assembly brick
2402.18695 (2024-02)
Grounding Language Models for Visual Entity Recognition
2406.10701 (2024-06)
MIND: Multimodal Shopping Intention Distillation from Large Vision-language Models for E-commerce Purchase Understanding
LlavaMetaModel
brick of this paper's assembly: helper
2 other papers
2 later · 1 only as an assembly brick
2406.20095 (2024-06)
LLaRA: Supercharging Robot Learning Data for Vision-Language Policy
2603.12930 (2026-03)
Rethinking VLMs for Image Forgery Detection and Localization
SimpleResBlock
brick of this paper's assembly: helper
2 other papers
2 later · 2 only as an assembly brick
2501.08282 (2025-01)
LLaVA-ST: A Multimodal Large Language Model for Fine-Grained Spatial-Temporal Understanding
2503.03803 (2025-03)
EgoLife: Towards Egocentric Life Assistant

Shown: the 20 functions present in the most other papers. 9 more shared functions are not listed here.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Instruction tuning large language models (LLMs) using machine-generated instruction-following data has been shown to improve zero-shot capabilities on new tasks, but the idea is less explored in the multimodal field. We present the first attempt to use language-only GPT-4 to generate multimodal language-image instruction-following data. By instruction tuning on such generated data, we introduce LLaVA: Large Language and Vision Assistant, an end-to-end trained large multimodal model that connects a vision encoder and an LLM for generalpurpose visual and language understanding. To facilitate future research on visual instruction following, we construct two evaluation benchmarks with diverse and challenging application-oriented tasks. Our experiments show that LLaVA demonstrates impressive multimodal chat abilities, sometimes exhibiting the behaviors of multimodal GPT-4 on unseen images/instructions, and yields a 85.1% relative score compared with GPT-4 on a synthetic multimodal instruction-following dataset. When fine-tuned on Science QA, the synergy of LLaVA and GPT-4 achieves a new state-of-the-art accuracy of 92.53%. We make GPT-4 generated visual instruction tuning data, our model, and code publicly available. Context type 1: Captions A group of people standing outside of a black vehicle with various luggage. Luggage surrounds a vehicle in an underground parking area People try to fit all of their luggage in an SUV. The sport utility vehicle is parked in the public garage, being packed for a trip Some people with luggage near a van that is transporting it.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2304.08485")
get_code_for_paper("2304.08485")
have("2304.08485")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2304.08485.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2304.08485.svg)](https://syntology.ai/paper/2304.08485)

Connect an agent — have() is free.