SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2212.09741 · 2022

One Embedder, Any Task: Instruction-Finetuned Text Embeddings

arXiv · PDF · Open in the Atlas

Code that ran

17 of 20 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 3 violated it and 7 have no contract verdict; 7 failures of our instrument; 3 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/2212.09741.json · badge: /api/ran/2212.09741.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
HKUNLP/instructor-embedding canonical (classifier label)
classifier basis not recorded on the link
15 ran · 3 no recorded run
xlang-ai/instructor-embedding canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
FunctionStatusWhere it lives
batch_to_device Ran
ran; our contract draft was wrong, not the code
xlang-ai/instructor-embedding/InstructorEmbedding/instructor.py
code served (permissive licence) · get_code("f567def389222696")
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 dict. 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 Tensor

torch.manual_seed(0)

# Construct a batch dict with various tensor types
batch = {
    "input_ids": torch.randint(0, 100, (2, 4)),
    "attention_mask": torch.ones((2, 4)),
    "token_type_ids": torch.zeros((2, 4)),
    "metadata": "not_a_tensor",
    "embeddings": torch.randn((2, 3, 4)),
}

target_device = "cpu"

ARGS = {
    "batch": batch,
    "target_device": target_device,
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
construct_prompt Ran
ran on a synthesized input
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_geo.py
code served (permissive licence) · get_code("012e0e03723d52ec")
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 str. 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

torch.manual_seed(0)

db_prompt = "SELECT * FROM users WHERE "
instructions = "Instructions: Write a SQL query to find the answer."
question = "What is the total number of active users?"

ARGS = {
    "db_prompt": db_prompt,
    "instructions": instructions,
    "question": question
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
conversion Ran
ran on a synthesized input
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_mwoz.py
code served (permissive licence) · get_code("5deafd882c22eab9")
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 str. 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

# Test case 1: normal conversion (forward direction)
prompt_1 = "I want to leaveat 10am and arriveby 5pm with food italian"
ARGS = {
    "prompt": prompt_1,
    "reverse": 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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
expand_to_aliases Ran
ran; our contract draft was wrong, not the code
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/utils.py
code served (permissive licence) · get_code("d51ed70aaef3b5d0")
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 set. 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 numpy as np

# Setup code to construct valid arguments for expand_to_aliases function

ARGS = {
    "given_answers": [
        "The_quick_brown_fox",
        "jumps-over, the 'fence'",
        "hello world",
        "test_answer-with.punctuation"
    ],
    "make_sub_answers": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type set — a set output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a set 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
fast_votek Ran
ran on a synthesized input
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/two_steps.py
code served (permissive licence) · get_code("5a28ddf2ee395e47")
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 list, 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 numpy as np
import torch

torch.manual_seed(0)
np.random.seed(0)

# Create small embeddings - shape (n_samples, embedding_dim)
# Using 6 samples with 4-dimensional embeddings
embeddings = np.random.randn(6, 4).astype(np.float32)

# Parameters for the function
select_num = 3  # Number of indices to select
k = 2           # Number of neighbors to consider in voting
vote_file = None  # No file I/O in sandbox

ARGS = {
    "embeddings": embeddings,
    "select_num": select_num,
    "k": k,
    "vote_file": vote_file
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, torch, tqdm — 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 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
find_indices_from_embeddings Ran
ran on a synthesized input
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_amazon.py
code served (permissive licence) · get_code("75fa1b0ab7064c3c")
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 list, 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 random
import torch
import numpy as np
from sklearn.metrics.pairwise import cosine_similarity

torch.manual_seed(0)
random.seed(0)

# Create small embeddings tensor: shape (5, 4) - 5 embeddings of dimension 4
embeddings = torch.randn(5, 4, dtype=torch.float32)

# Parameters
select_num = 3
mean_normal = False

ARGS = {
    "embeddings": embeddings,
    "select_num": select_num,
    "mean_normal": mean_normal,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, 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 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
format_dataset Ran
ran; our contract draft was wrong, not the code
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/get_task.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from cmavro/packllm, is pointer only (licence: NONE) · get_code("82fab1f6e6bddec2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from cmavro/packllm, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

get_repeat_num Ran
ran on a synthesized input
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_mwoz.py
code served (permissive licence) · get_code("d8b8be7ec3cf207f")
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 int, ndim 0, dtype family int, 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 os
import tempfile
import shutil

# Create a temporary cache directory with test subdirectories
temp_cache_dir = tempfile.mkdtemp(prefix='cache_test_')
original_cwd = os.getcwd()
os.chdir(temp_cache_dir)

# Create cache subdirectory
os.makedirs('cache', exist_ok=True)

# Create some test directories with naming patterns
os.makedirs('cache/experiment_1', exist_ok=True)
os.makedirs('cache/experiment_2', exist_ok=True)
os.makedirs('cache/model_run_5', exist_ok=True)
os.makedirs('cache/other_data_10', exist_ok=True)

ARGS = {
    'prefix': None
}
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 int, ndim 0, dtype family int, 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
get_sub_answers Ran
ran; our contract draft was wrong, not the code
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/utils.py
code served (permissive licence) · get_code("91631e556572af73")
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 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

# Analyze the function:
# - answers: a list that is iterated over
# - each element x in answers is a string (called with .split(' '))
# - begin: int, default 0
# - end: int or None, default None
# The function filters answers where len(x.split(' ')) > 1 (multi-word strings)
# and returns sliced joined strings

ARGS = {
    "answers": [
        "hello world",
        "foo bar baz",
        "single",
        "one two three four",
        "alpha beta"
    ],
    "begin": 0,
    "end": None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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
import_from_string Ran
ran; our contract draft was wrong, not the code
xlang-ai/instructor-embedding/InstructorEmbedding/instructor.py
code served (permissive licence) · get_code("fc1006b014fb9c69")
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 type. 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 importlib
import torch

torch.manual_seed(0)

# The function import_from_string takes a single parameter: dotted_path
# It expects a string in the format "module.path.ClassName" or "module.path.function_name"
# The function will:
# 1. Split the dotted_path by the last dot to get module_path and class_name
# 2. Try to import the full dotted_path as a module first
# 3. If that fails, import the module_path
# 4. Get the attribute class_name from the module

# Valid test cases:
# - A built-in like "os.path" would give us the path module
# - A class like "torch.nn.Linear" would give us the Linear class
# - A function like "os.path.join" would give us the join function

ARGS = {
    "dotted_path": "torch.nn.Linear"
}
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 type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
is_commonword Ran
ran, violated the contract we drafted
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/bridge_content_encoder.py
code served (permissive licence) · get_code("b724cc2d4066a429")
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 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)

# The function is_commonword takes a single string parameter 's'
# and checks if its stripped version is in the _commonwords set
# No complex objects needed - just a simple string

ARGS = {
    's': 'yes'
}
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 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
is_number Ran
ran, violated the contract we drafted
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/bridge_content_encoder.py
code served (permissive licence) · get_code("f39fdc1ce6cffe08")
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 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 numpy as np

torch.manual_seed(0)

# The function is_number takes a single string parameter 's'
# It tries to convert the string to a float after removing commas
# We need to provide valid test cases

ARGS = {
    "s": "123.45"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, 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
is_stopword Ran
ran, violated the contract we drafted
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/bridge_content_encoder.py
code served (permissive licence) · get_code("d56a1f9221823a48")
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 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 numpy as np

# The function is_stopword takes a single parameter: s (a string)
# It checks if s.strip() is in the _stopwords set
# So we need to provide a string argument

ARGS = {
    "s": "hello"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, 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
maybe_add_quotes Ran
ran on a synthesized input
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_geo.py
code served (permissive licence) · get_code("678526b83af152ab")
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 str. 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

def maybe_add_quotes(val):
    if isinstance(val, str):
        return "'" + val + "'"
    return str(val)

# Test cases for the function
ARGS = {
    "val": "hello"
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
mean_pooling Ran
ran on a synthesized input
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_coda_title_gen.py
code served (permissive licence) · get_code("597e15632d2a22fb")
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 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

torch.manual_seed(0)

# Construct tiny tensors for mean_pooling
batch_size = 2
seq_length = 3
embedding_dim = 4

# model_output: token embeddings of shape (batch_size, seq_length, embedding_dim)
model_output = torch.randn(batch_size, seq_length, embedding_dim)

# attention_mask: binary mask of shape (batch_size, seq_length)
attention_mask = torch.tensor([
    [1, 1, 0],
    [1, 1, 1]
], dtype=torch.long)

ARGS = {
    "model_output": model_output,
    "attention_mask": attention_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 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: none by the harness; any seed is in the setup block
process_mnli_examples Ran
ran; our contract draft was wrong, not the code
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/get_task.py
code served (permissive licence) · get_code("071f4dd2f4d4b22c")
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 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
from tqdm import tqdm

def process_mnli_examples(examples):
    processed_examples = []
    idx = 0
    for raw_data in tqdm(examples, desc='process mnli examples'):
        processed_examples.append({'id': idx, 'label': raw_data['label'], 'premise': raw_data['premise'], 'hypothesis': raw_data['hypothesis']})
        idx += 1
    return processed_examples

# Create example MNLI data
examples = [
    {
        'label': 0,
        'premise': 'A man is walking in the park.',
        'hypothesis': 'Someone is outside.'
    },
    {
        'label': 1,
        'premise': 'The cat is sleeping on the couch.',
        'hypothesis': 'The dog is barking loudly.'
    }
]

ARGS = {
    'examples': examples
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm — 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
process_rte_examples Ran
ran; our contract draft was wrong, not the code
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/get_task.py
code served (permissive licence) · get_code("b775e4262bf408f6")
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 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
from tqdm import tqdm

# Construct example data matching what process_rte_examples expects
examples = [
    {
        'id': 0,
        'label': 0,
        'sentence1': 'The cat sat on the mat.',
        'sentence2': 'A feline rested on a rug.'
    },
    {
        'id': 1,
        'label': 1,
        'sentence1': 'It is raining outside.',
        'sentence2': 'The weather is clear and sunny.'
    }
]

ARGS = {
    'examples': examples
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm — 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
postprocess_text Not yet run
unverified (harvested, no recorded run)
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_coda_title_gen.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from inhwanbae/LMTrajectory, is pointer only (licence: NOASSERTION) · get_code("eff19094fe0ec70e")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from inhwanbae/LMTrajectory, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

postprocess_text Not yet run
unverified (harvested, no recorded run)
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_answerbility.py
code served (permissive licence) · get_code("8d5099dd630ccb92")
How it ran

no invocation recorded.

v2_vote_k_select Not yet run
unverified (harvested, no recorded run)
HKUNLP/instructor-embedding/evaluation/prompt_retrieval/main_mwoz.py
code served (permissive licence) · get_code("4db411255e0de092")
How it ran

no invocation recorded.

Also present in other papers' code

Of 20 samples lifted from this paper's linked repositories, 20 carry a function identity and 0 do not. Of the 20 distinct functions with an identity, 10 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 9 later) and 10 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
is_commonword 5 other papers
1 dated earlier, 4 later
2109.05093 (2021-09)
PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models
2306.08891 (2023-06)
Interleaving Pre-Trained Language Models and Large Language Models for Zero-Shot NL2SQL Generation
2310.13659 (2023-10)
Benchmarking and Improving Text-to-SQL Generation under Ambiguity
and 2 more not named here
is_number 5 other papers
1 dated earlier, 4 later
2109.05093 (2021-09)
PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models
2306.08891 (2023-06)
Interleaving Pre-Trained Language Models and Large Language Models for Zero-Shot NL2SQL Generation
2310.13659 (2023-10)
Benchmarking and Improving Text-to-SQL Generation under Ambiguity
and 2 more not named here
is_stopword 5 other papers
1 dated earlier, 4 later
2109.05093 (2021-09)
PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models
2306.08891 (2023-06)
Interleaving Pre-Trained Language Models and Large Language Models for Zero-Shot NL2SQL Generation
2310.13659 (2023-10)
Benchmarking and Improving Text-to-SQL Generation under Ambiguity
and 2 more not named here
format_dataset 2 other papers
2 later
2310.20046 (2023-10)
Which Examples to Annotate for In-Context Learning? Towards Effective and Efficient Selection
2404.11531 (2024-04)
Pack of LLMs: Model Fusion at Test-Time via Perplexity Optimization
postprocess_text 2 other papers
2 later
2310.20046 (2023-10)
Which Examples to Annotate for In-Context Learning? Towards Effective and Efficient Selection
2403.18447 (2024-03)
Can Language Beat Numerical Regression? Language-Based Multimodal Trajectory Prediction
conversion 1 other paper
1 same year, order not determinable
2022.findings-emnlp.193 (2022 (year only))
In-Context Learning for Few-Shot Dialogue State Tracking
expand_to_aliases 1 other paper
1 later
2310.20046 (2023-10)
Which Examples to Annotate for In-Context Learning? Towards Effective and Efficient Selection
get_sub_answers 1 other paper
1 later
2310.20046 (2023-10)
Which Examples to Annotate for In-Context Learning? Towards Effective and Efficient Selection
process_mnli_examples 1 other paper
1 later
2310.20046 (2023-10)
Which Examples to Annotate for In-Context Learning? Towards Effective and Efficient Selection
process_rte_examples 1 other paper
1 later
2310.20046 (2023-10)
Which Examples to Annotate for In-Context Learning? Towards Effective and Efficient Selection

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

We introduce INSTRUCTOR, a new method for computing text embeddings given task instructions: every text input is embedded together with instructions explaining the use case (e.g., task and domain descriptions). Unlike encoders from prior work that are more specialized, INSTRUCTOR is a single embedder that can generate text embeddings tailored to different downstream tasks and domains, without any further training. We first annotate instructions for 330 diverse tasks and train INSTRUCTOR on this multitask mixture with a contrastive loss. We evaluate INSTRUCTOR on 70 embedding evaluation tasks (66 of which are unseen during training), ranging from classification and information retrieval to semantic textual similarity and text generation evaluation. INSTRUCTOR, while having an order of magnitude fewer parameters than the previous best model, achieves state-of-the-art performance, with an average improvement of 3.4% compared to the previous best results on the 70 diverse datasets. Our analysis suggests that INSTRUCTOR is robust to changes in instructions, and that instruction finetuning mitigates the challenge of training a single model on diverse datasets. Our model, code, and data are available at https://instructor-embedding.github.io.

For agents

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

get_harvested_code_for_paper("2212.09741")
get_code_for_paper("2212.09741")
have("2212.09741")

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

curl https://syntology.ai/api/ran/2212.09741.json

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

[![Syntology run record](https://syntology.ai/api/ran/2212.09741.svg)](https://syntology.ai/paper/2212.09741)

Connect an agent — have() is free.