SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.08409 · CVPR · 2023

LANA: A Language-Capable Navigator for Instruction Following and Generation

Yi Yang, Wenguan Wang, Xiaohan Wang, Jiayi Shao

arXiv · PDF · Open in the Atlas

Code that ran

11 of 15 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (8 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 11 have no contract verdict; 0 failures of our instrument; 4 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/2303.08409.json · badge: /api/ran/2303.08409.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
wxh1996/lana-vln marked official in Papers with Code
Papers with Code archive
11 ran · 4 no recorded run
FunctionStatusWhere it lives
ActionEmbeddings Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("95201a35b73fcc8b")
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, ActionEmbeddings: 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
from torch import nn

torch.manual_seed(0)

class Config:
    def __init__(self):
        self.image_feat_size = 3
        self.angle_feat_size = 5
        self.hidden_size = 4
        self.hidden_dropout_prob = 0.1

config = Config()

# Create tiny tensors for forward pass
batch_size = 2
img_feat = torch.randn(batch_size, config.image_feat_size)
ang_feat = torch.randn(batch_size, config.angle_feat_size)
type_embeddings = torch.randn(batch_size, 1, config.hidden_size)
step_embedding = torch.randn(batch_size, config.hidden_size)

ARGS = {
    "config": config,
}
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 ActionEmbeddings — 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 ActionEmbeddings 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
BertAttention Ran
ran on a synthesized input
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("972bc2726cd67bc8")
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 tuple. 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 math
from torch import nn

class Config:
    def __init__(self):
        self.hidden_size = 8
        self.num_attention_heads = 2
        self.attention_probs_dropout_prob = 0.1
        self.hidden_dropout_prob = 0.1
        self.layer_norm_eps = 1e-12
        self.output_attentions = False

config = Config()

INIT_ARGS = {
    'config': config
}

batch_size = 2
seq_length = 5
hidden_size = config.hidden_size

input_tensor = torch.randn(batch_size, seq_length, hidden_size)
attention_mask = torch.zeros(batch_size, 1, 1, seq_length)

ARGS = {
    'input_tensor': input_tensor,
    'attention_mask': attention_mask,
    'head_mask': None,
    'future_mask': None
}
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 tuple — a tuple output has no array shape of its own; none is 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
BertLMPredictionHead Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("46726083e44b793d")
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, BertLMPredictionHead: 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
import math

torch.manual_seed(0)

def gelu(x):
    """Implementation of the gelu activation function."""
    return x * 0.5 * (1.0 + torch.erf(x / math.sqrt(2.0)))

def swish(x):
    return x * torch.sigmoid(x)

ACT2FN = {'gelu': gelu, 'relu': torch.nn.functional.relu, 'swish': swish}

BertLayerNorm = torch.nn.LayerNorm

class BertPredictionHeadTransform(nn.Module):

    def __init__(self, config):
        super(BertPredictionHeadTransform, self).__init__()
        clip_size = 512
        self.dense = nn.Linear(config.hidden_size, clip_size)
        self.LayerNorm = BertLayerNorm(clip_size, eps=config.layer_norm_eps)
        if isinstance(config.hidden_act, str):
            self.transform_act_fn = ACT2FN[config.hidden_act]
        else:
            self.transform_act_fn = config.hidden_act

    def forward(self, hidden_states):
        hidden_states = self.dense(hidden_states)
        hidden_states = self.transform_act_fn(hidden_states)
        hidden_states = self.LayerNorm(hidden_states)
        return hidden_states

class BertLMPredictionHead(nn.Module):

    def __init__(self, config):
        super(BertLMPredictionHead, self).__init__()
        self.transform = BertPredictionHeadTransform(config)
        clip_size = 512
        self.decoder = nn.Linear(clip_size, config.vocab_size, bias=False)
        self.bias = nn.Parameter(torch.zeros(config.vocab_size))

    def forward(self, hidden_states):
        hidden_states = self.transform(hidden_states)
        hidden_states = self.decoder(hidden_states) + self.bias
        return hidden_states

class Config:
    def __init__(self):
        self.hidden_size = 8
        self.vocab_size = 16
        self.layer_norm_eps = 1e-12
        self.hidden_act = 'gelu'

config = Config()

ARGS = {
    'config': config
}
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 BertLMPredictionHead — 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 BertLMPredictionHead 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
BertOnlyMLMHead Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("8439e6da38dfae41")
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, BertOnlyMLMHead: 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
import math

torch.manual_seed(0)

def gelu(x):
    return x * 0.5 * (1.0 + torch.erf(x / math.sqrt(2.0)))

def swish(x):
    return x * torch.sigmoid(x)

ACT2FN = {'gelu': gelu, 'relu': torch.nn.functional.relu, 'swish': swish}

BertLayerNorm = torch.nn.LayerNorm

class BertPredictionHeadTransform(nn.Module):
    def __init__(self, config):
        super(BertPredictionHeadTransform, self).__init__()
        clip_size = 512
        self.dense = nn.Linear(config.hidden_size, clip_size)
        self.LayerNorm = BertLayerNorm(clip_size, eps=config.layer_norm_eps)
        if isinstance(config.hidden_act, str):
            self.transform_act_fn = ACT2FN[config.hidden_act]
        else:
            self.transform_act_fn = config.hidden_act

    def forward(self, hidden_states):
        hidden_states = self.dense(hidden_states)
        hidden_states = self.transform_act_fn(hidden_states)
        hidden_states = self.LayerNorm(hidden_states)
        return hidden_states

class BertLMPredictionHead(nn.Module):
    def __init__(self, config):
        super(BertLMPredictionHead, self).__init__()
        self.transform = BertPredictionHeadTransform(config)
        clip_size = 512
        self.decoder = nn.Linear(clip_size, config.vocab_size, bias=False)
        self.bias = nn.Parameter(torch.zeros(config.vocab_size))

    def forward(self, hidden_states):
        hidden_states = self.transform(hidden_states)
        hidden_states = self.decoder(hidden_states) + self.bias
        return hidden_states

class BertOnlyMLMHead(nn.Module):
    def __init__(self, config):
        super(BertOnlyMLMHead, self).__init__()
        self.predictions = BertLMPredictionHead(config)

    def forward(self, sequence_output):
        prediction_scores = self.predictions(sequence_output)
        return prediction_scores

class Config:
    def __init__(self):
        self.hidden_size = 16
        self.vocab_size = 8
        self.layer_norm_eps = 1e-12
        self.hidden_act = 'gelu'

config = Config()

ARGS = {
    'config': config
}
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 BertOnlyMLMHead — 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 BertOnlyMLMHead 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
BertPredictionHeadTransform Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("fbfefa916c6de2bb")
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, BertPredictionHeadTransform: 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 torch import nn as nn_module

torch.manual_seed(0)

class Config:
    def __init__(self):
        self.hidden_size = 4
        self.layer_norm_eps = 1e-12
        self.hidden_act = 'gelu'

config = Config()

ARGS = {
    'config': config
}
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 BertPredictionHeadTransform — 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 BertPredictionHeadTransform 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
BertSelfAttention Ran
ran on a synthesized input
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("a0efcc75e5e7e5de")
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 tuple. 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 math
import torch
from torch import nn

class BertSelfAttention(nn.Module):

    def __init__(self, config):
        super(BertSelfAttention, self).__init__()
        if config.hidden_size % config.num_attention_heads != 0:
            raise ValueError('The hidden size (%d) is not a multiple of the number of attention heads (%d)' % (config.hidden_size, config.num_attention_heads))
        self.output_attentions = config.output_attentions
        self.num_attention_heads = config.num_attention_heads
        self.attention_head_size = int(config.hidden_size / config.num_attention_heads)
        self.all_head_size = self.num_attention_heads * self.attention_head_size
        self.query = nn.Linear(config.hidden_size, self.all_head_size)
        self.key = nn.Linear(config.hidden_size, self.all_head_size)
        self.value = nn.Linear(config.hidden_size, self.all_head_size)
        self.dropout = nn.Dropout(config.attention_probs_dropout_prob)

    def transpose_for_scores(self, x):
        new_x_shape = x.size()[:-1] + (self.num_attention_heads, self.attention_head_size)
        x = x.view(*new_x_shape)
        return x.permute(0, 2, 1, 3)

    def forward(self, hidden_states, attention_mask, head_mask=None, future_mask=None):
        mixed_query_layer = self.query(hidden_states)
        mixed_key_layer = self.key(hidden_states)
        mixed_value_layer = self.value(hidden_states)
        query_layer = self.transpose_for_scores(mixed_query_layer)
        key_layer = self.transpose_for_scores(mixed_key_layer)
        value_layer = self.transpose_for_scores(mixed_value_layer)
        attention_scores = torch.matmul(query_layer, key_layer.transpose(-1, -2))
        attention_scores = attention_scores / math.sqrt(self.attention_head_size)
        attention_scores = attention_scores + attention_mask
        if future_mask is not None:
            attention_scores = attention_scores.masked_fill(future_mask.unsqueeze(0).unsqueeze(1) != 0, float('-inf'))
        attention_probs = nn.Softmax(dim=-1)(attention_scores)
        attention_probs = self.dropout(attention_probs)
        if head_mask is not None:
            attention_probs = attention_probs * head_mask
        context_layer = torch.matmul(attention_probs, value_layer)
        context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
        new_context_layer_shape = context_layer.size()[:-2] + (self.all_head_size,)
        context_layer = context_layer.view(*new_context_layer_shape)
        outputs = (context_layer, attention_scores) if self.output_attentions else (context_layer,)
        return outputs


class Config:
    def __init__(self, hidden_size, num_attention_heads, attention_probs_dropout_prob, output_attentions):
        self.hidden_size = hidden_size
        self.num_attention_heads = num_attention_heads
        self.attention_probs_dropout_prob = attention_probs_dropout_prob
        self.output_attentions = output_attentions


INIT_ARGS = {
    'config': Config(
        hidden_size=8,
        num_attention_heads=2,
        attention_probs_dropout_prob=0.1,
        output_attentions=True
    )
}

ARGS = {
    'hidden_states': torch.randn(2, 3, 8),
    'attention_mask': torch.zeros(2, 1, 1, 3),
    'head_mask': None,
    'future_mask': None
}
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 tuple — a tuple output has no array shape of its own; none is 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
InputSelfAttention Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("ab4507b7278676b1")
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, InputSelfAttention: 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
from torch import nn

torch.manual_seed(0)

class Config:
    def __init__(self):
        self.hidden_size = 4
        self.num_attention_heads = 2
        self.attention_probs_dropout_prob = 0.1
        self.hidden_dropout_prob = 0.1
        self.layer_norm_eps = 1e-12

config = Config()

batch_size = 2
seq_length = 3

input_tensor = torch.randn(batch_size, seq_length, config.hidden_size)
input_mask = torch.ones(batch_size, 1, 1, seq_length)

ARGS = {
    "config": config,
    "ctx_dim": None,
}
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 InputSelfAttention — 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 InputSelfAttention 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
InputXAttention Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("10473dc20e4ea297")
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, InputXAttention: 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
from torch import nn

torch.manual_seed(0)

class Config:
    def __init__(self):
        self.hidden_size = 4
        self.num_attention_heads = 2
        self.attention_probs_dropout_prob = 0.1
        self.hidden_dropout_prob = 0.1
        self.layer_norm_eps = 1e-12

config = Config()

batch_size = 2
seq_len_input = 3
seq_len_ctx = 5
hidden_size = 4

input_tensor = torch.randn(batch_size, seq_len_input, hidden_size)
ctx_tensor = torch.randn(batch_size, seq_len_ctx, hidden_size)
input_mask = torch.ones(batch_size, 1, 1, seq_len_input)
ctx_att_mask = torch.ones(batch_size, 1, 1, seq_len_ctx)

ARGS = {
    "config": config,
    "ctx_dim": None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
ipdb, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type InputXAttention — 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 InputXAttention 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
ResidualAttentionBlock Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("e60a700608513530")
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, ResidualAttentionBlock: 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
from torch import nn
from collections import OrderedDict

torch.manual_seed(0)

class LayerNorm(nn.LayerNorm):
    """Subclass torch's LayerNorm to handle fp16."""

    def forward(self, x: torch.Tensor):
        orig_type = x.dtype
        ret = super().forward(x.type(torch.float32))
        return ret.type(orig_type)

class QuickGELU(nn.Module):

    def forward(self, x: torch.Tensor):
        return x * torch.sigmoid(1.702 * x)

class ResidualAttentionBlock(nn.Module):

    def __init__(self, d_model: int, n_head: int, attn_mask: torch.Tensor=None):
        super().__init__()
        self.attn = nn.MultiheadAttention(d_model, n_head)
        self.ln_1 = LayerNorm(d_model)
        self.mlp = nn.Sequential(OrderedDict([('c_fc', nn.Linear(d_model, d_model * 4)), ('gelu', QuickGELU()), ('c_proj', nn.Linear(d_model * 4, d_model))]))
        self.ln_2 = LayerNorm(d_model)
        self.attn_mask = attn_mask

    def attention(self, x: torch.Tensor):
        attn_mask_ = self.attn_mask
        if self.attn_mask is not None and hasattr(self.attn_mask, '__call__'):
            attn_mask_ = self.attn_mask(x.size(0))
        attn_mask_ = attn_mask_.to(dtype=x.dtype, device=x.device) if attn_mask_ is not None else None
        return self.attn(x, x, x, need_weights=False, attn_mask=attn_mask_)[0]

    def forward(self, x: torch.Tensor):
        x = x + self.attention(self.ln_1(x))
        x = x + self.mlp(self.ln_2(x))
        return x

# Create a simple callable attn_mask that returns a mask tensor
def make_attn_mask(seq_len):
    return torch.zeros((seq_len, seq_len), dtype=torch.float32)

ARGS = {
    'd_model': 4,
    'n_head': 2,
    'attn_mask': make_attn_mask
}
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 ResidualAttentionBlock — 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 ResidualAttentionBlock 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
Transformer Ran
ran on a synthesized input · constructed an object; no result was computed
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("034d394f08c70794")
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, Transformer: 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
from torch import nn

torch.manual_seed(0)

# Parameters for Transformer
width = 8  # embedding dimension
layers = 2  # number of transformer layers
heads = 2  # number of attention heads
attn_mask = None  # no attention mask (optional parameter)

ARGS = {
    'width': width,
    'layers': layers,
    'heads': heads,
    'attn_mask': attn_mask
}
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 Transformer — 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 Transformer 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
VisualEmbeddings Ran
ran on a synthesized input
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("8515b5135628725d")
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 tuple. 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
from torch import nn

BertLayerNorm = torch.nn.LayerNorm

class VisualEmbeddings(nn.Module):

    def __init__(self, config):
        super().__init__()
        self.cls_token = nn.Parameter(torch.zeros(1, 1, config.hidden_size))
        self.pano_img_linear = nn.Linear(config.image_feat_size, config.hidden_size)
        self.pano_img_layer_norm = BertLayerNorm(config.hidden_size, eps=1e-12)
        self.pano_ang_linear = nn.Linear(config.angle_feat_size, config.hidden_size)
        self.pano_ang_layer_norm = BertLayerNorm(config.hidden_size, eps=1e-12)
        self.type_embedding = nn.Embedding(1, config.hidden_size)
        self.layer_norm = BertLayerNorm(config.hidden_size, eps=1e-12)
        self.dropout = nn.Dropout(config.hidden_dropout_prob)

    @property
    def device(self):
        return self.pano_img_linear.weight.device

    def forward(self, pano_img_feats, pano_ang_feats, step_embedding=None, batch_size=None):
        type_ids = torch.zeros((batch_size,)).long().to(self.device)
        type_embeddings = self.type_embedding(type_ids).unsqueeze(1)
        cls_embeddings = self.dropout(self.layer_norm(self.cls_token.expand(batch_size, -1, -1) + type_embeddings.squeeze(1)))
        if pano_img_feats is not None:
            batch_size, num_pano, _ = pano_img_feats.size()
            pano_embeddings = self.pano_img_layer_norm(self.pano_img_linear(pano_img_feats)) + self.pano_ang_layer_norm(self.pano_ang_linear(pano_ang_feats)) + type_embeddings
            pano_embeddings = pano_embeddings + step_embedding
            pano_embeddings = self.layer_norm(pano_embeddings)
            pano_embeddings = self.dropout(pano_embeddings)
        else:
            pano_embeddings = None
        return (cls_embeddings, pano_embeddings)


class Config:
    def __init__(self):
        self.hidden_size = 8
        self.image_feat_size = 4
        self.angle_feat_size = 3
        self.hidden_dropout_prob = 0.1


INIT_ARGS = {
    'config': Config()
}

ARGS = {
    'pano_img_feats': torch.randn(2, 5, 4),
    'pano_ang_feats': torch.randn(2, 5, 3),
    'step_embedding': torch.randn(2, 5, 8),
    'batch_size': 2
}

if __name__ == '__main__':
    model = VisualEmbeddings(**INIT_ARGS)
    cls_emb, pano_emb = model(**ARGS)
    print("cls_embeddings shape:", cls_emb.shape)
    print("pano_embeddings shape:", pano_emb.shape if pano_emb is not None else None)
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 tuple — a tuple output has no array shape of its own; none is 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
BertLayer Not yet run
unverified (harvested, no recorded run)
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("71b64ba0c2e09916")
How it ran

no invocation recorded.

LXRTXLayer Not yet run
unverified (harvested, no recorded run)
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("1aa0b58638317139")
How it ran

no invocation recorded.

LxmertEncoder Not yet run
unverified (harvested, no recorded run)
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("c7aa8c5278bebd66")
How it ran

no invocation recorded.

NavCMT Not yet run
unverified (harvested, no recorded run)
wxh1996/lana-vln/finetune_src/models/vilmodel_cmt_lana.py
code served (permissive licence) · get_code("c8a9c4bd19048e94")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 3 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 1 the same month, 3 later) and 12 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
Transformer
brick of this paper's assembly: helper
16 other papers
3 dated earlier, 1 the same month, 12 later · 10 only as an assembly brick
2105.01879 (2021-05)
MOS: Towards Scaling Out-of-distribution Detection for Large Semantic Space
2205.14459 (2022-05)
CYCLIP: Cyclic Contrastive Language-Image Pretraining
2301.06958 (2023-01)
RILS: Masked Visual Reconstruction in Language Semantic Space
and 13 more not named here
BertLayer
brick of this paper's assembly: model
2 other papers
1 dated earlier, 1 later
2110.13309 (2021-10)
History Aware Multimodal Transformer for Vision-and-Language Navigation
2308.12587 (2023-08)
Grounded Entity-Landmark Adaptive Pre-training for Vision-and-Language Navigation
BertOnlyMLMHead
brick of this paper's assembly: helper
1 other paper
1 later
2308.12587 (2023-08)
Grounded Entity-Landmark Adaptive Pre-training for Vision-and-Language Navigation

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

Recently, visual-language navigation (VLN) -entailing robot agents to follow navigation instructions -has shown great advance. However, existing literature put most emphasis on interpreting instructions into actions, only delivering "dumb" wayfinding agents. In this article, we devise LANA, a language-capable navigation agent which is able to not only execute human-written navigation commands, but also provide route descriptions to humans. This is achieved by simultaneously learning instruction following and generation with only one single model. More specifically, two encoders, respectively for route and language encoding, are built and shared by two decoders, respectively for action prediction and instruction generation, so as to exploit cross-task knowledge and capture task-specific characteristics. Throughout pretraining and fine-tuning, both instruction following and generation are set as optimization objectives. We empirically verify that, compared with recent advanced task-specific solutions, LANA attains better performances on both instruction following and route description, with nearly half complexity. In addition, endowed with language generation capability, LANA can explain to human its behaviours and assist human's wayfinding. This work is expected to foster future efforts towards building more trustworthy and sociallyintelligent navigation robots.

For agents

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

get_harvested_code_for_paper("2303.08409")
get_code_for_paper("2303.08409")
have("2303.08409")

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

curl https://syntology.ai/api/ran/2303.08409.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.08409.svg)](https://syntology.ai/paper/2303.08409)

Connect an agent — have() is free.