SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.13773 · 2026

On Representation Redundancy in Large-Scale Instruction Tuning Data Selection

Jun Zhou, Wenjie Qu, Ziyao Guo, Youwei Shu, Dingnan Jin, Jiaheng Zhang, Qing Cui, Shaomian Zheng

arXiv · PDF · Open in the Atlas

Code that ran

9 of 14 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, 0 violated it and 1 has no contract verdict; 8 failures of our instrument; 5 no recorded run), measured 2026-11-22; 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/2602.13773.json · badge: /api/ran/2602.13773.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
tdano1/CRDS canonical (classifier label)
classifier basis: census
9 ran · 5 no recorded run
FunctionStatusWhere it lives
construct_test_sample Ran
ran; our contract draft was wrong, not the code
tdano1/CRDS/main/data.py
code served (permissive licence) · get_code("68773c10680996fa")
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 transformers import AutoTokenizer

torch.manual_seed(0)

# Create a minimal tokenizer stand-in
class MinimalTokenizer:
    def __init__(self):
        self.eos_token = "</s>"
        self.bos_token = "<s>"
        self.pad_token_id = 0
    
    def __call__(self, text, return_tensors=None, max_length=2048, truncation=False, add_special_tokens=True):
        # Simple tokenization: split and convert to token ids
        tokens = text.split()
        token_ids = [hash(t) % 1000 for t in tokens]  # Simple hash-based tokenization
        
        if max_length and truncation and len(token_ids) > max_length:
            token_ids = token_ids[:max_length]
        
        token_ids_tensor = torch.tensor([token_ids], dtype=torch.long)
        attention_mask = torch.ones_like(token_ids_tensor)
        
        if return_tensors == 'pt':
            return {
                'input_ids': token_ids_tensor,
                'attention_mask': attention_mask
            }
        
        return {
            'input_ids': token_ids_tensor,
            'attention_mask': attention_mask
        }

# Create sample data
tokenizer = MinimalTokenizer()

sample = {
    'prompts': 'What is the capital of France?\n<|assistant|>\n',
    'labels': 'Paris is the capital of France.'
}

ARGS = {
    'tokenizer': tokenizer,
    'sample': sample,
    'max_length': 2048,
    'prompt_only': False,
    'response_only': 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 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
create_prompt_with_tulu_chat_format Ran
ran; our contract draft was wrong, not the code
tdano1/CRDS/main/utils.py
code served (permissive licence) · get_code("d0e89feb8d971deb")
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
from transformers import AutoTokenizer

# Create a minimal tokenizer stand-in with required attributes
class MinimalTokenizer:
    def __init__(self):
        self.eos_token = '</s>'
        self.bos_token = '<s>'

# Create messages in the Tulu chat format
messages = [
    {'role': 'system', 'content': 'You are a helpful assistant.'},
    {'role': 'user', 'content': 'What is 2+2?'},
    {'role': 'assistant', 'content': 'The answer is 4.'},
]

# Create tokenizer
tokenizer = MinimalTokenizer()

# Define all arguments
ARGS = {
    'messages': messages,
    'tokenizer': tokenizer,
    'add_bos': True,
    'prompt_only': False,
    'response_only': False,
    'no_special_tokens': 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 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
encode_with_messages_format_ling Ran
ran; our contract draft was wrong, not the code
tdano1/CRDS/main/utils.py
code served (permissive licence) · get_code("37b8ad3633ce227a")
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 transformers import AutoTokenizer

torch.manual_seed(0)

# Create a minimal tokenizer stand-in that has the required eos_token attribute
class MinimalTokenizer:
    def __init__(self):
        self.eos_token = "<|end|>"
    
    def __call__(self, text, return_tensors=None, max_length=None, truncation=False):
        # Simple mock tokenization: split by spaces and convert to token ids
        tokens = text.split()
        token_ids = [hash(token) % 1000 for token in tokens]  # Generate fake token ids
        
        if max_length and len(token_ids) > max_length:
            token_ids = token_ids[:max_length]
        
        input_ids = torch.tensor([token_ids], dtype=torch.long)
        attention_mask = torch.ones_like(input_ids)
        
        class TokenizerOutput:
            pass
        
        output = TokenizerOutput()
        output.input_ids = input_ids
        output.attention_mask = attention_mask
        return output

# Create example data
example = {
    'system_message': 'You are a helpful assistant.',
    'messages': [
        {'role': 'HUMAN', 'content': 'Hello, how are you?'},
        {'role': 'ASSISTANT', 'content': 'I am doing well, thank you for asking.'},
        {'role': 'OBSERVATION', 'content': 'This is an observation.'},
        {'role': 'HUMAN', 'content': 'What is the weather?'}
    ]
}

tokenizer = MinimalTokenizer()
max_seq_length = 128

ARGS = {
    'example': example,
    'tokenizer': tokenizer,
    'max_seq_length': max_seq_length
}
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 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
estimate_pass_at_k Ran
ran; our contract draft was wrong, not the code
tdano1/CRDS/main/eval/codex_humaneval/evaluation.py
code served (permissive licence) · get_code("ff29b26b4a0b7f51")
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 ndarray, ndim 1, dtype family float, shape (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 numpy as np

ARGS = {
    "num_samples": 100,
    "num_correct": np.array([10, 25, 50, 75]),
    "k": 5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (4,) — shape (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
generate_completions Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
tdano1/CRDS/main/eval/utils.py
code served (permissive licence) · get_code("cdbcc01805fac313")
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. 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 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 torch.nn as nn
from transformers import PreTrainedModel, PreTrainedTokenizer

torch.manual_seed(0)

# Minimal mock tokenizer
class MockTokenizer:
    def __init__(self):
        self.eos_token_id = 2
        self.pad_token_id = 0
        self.bos_token_id = 1
    
    def __call__(self, texts, padding=None, return_tensors=None, add_special_tokens=True):
        # Return a mock tokenization result
        batch_size = len(texts) if isinstance(texts, list) else 1
        seq_len = 5
        
        class TokenizationResult:
            def __init__(self):
                self.input_ids = torch.randint(3, 10, (batch_size, seq_len))
                self.attention_mask = torch.ones(batch_size, seq_len)
        
        return TokenizationResult()
    
    def batch_decode(self, token_ids, skip_special_tokens=True):
        # Return mock decoded strings
        batch_size = token_ids.shape[0]
        return ["mock output " + str(i) for i in range(batch_size)]

# Minimal mock model
class MockModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 4)
        self.device_obj = torch.device('cpu')
    
    @property
    def device(self):
        return self.device_obj
    
    def generate(self, input_ids, attention_mask, eos_token_id, stopping_criteria=None, **kwargs):
        batch_size = input_ids.shape[0]
        num_return_sequences = kwargs.get('num_return_sequences', 1)
        # Generate mock output tokens
        output_length = input_ids.shape[1] + 3
        outputs = torch.randint(3, 10, (batch_size * num_return_sequences, output_length))
        return outputs

model = MockModel()
tokenizer = MockTokenizer()

prompts = [
    "This is a test prompt.",
    "Another test prompt."
]

stop_id_sequences = [[2], [0, 2]]

ARGS = {
    'model': model,
    'tokenizer': tokenizer,
    'prompts': prompts,
    'batch_size': 2,
    'stop_id_sequences': stop_id_sequences,
    'add_special_tokens': True,
    'disable_tqdm': True,
    'max_new_tokens': 10,
    'num_return_sequences': 1,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm, transformers — 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_next_word_predictions Ran
ran; our contract draft was wrong, not the code
tdano1/CRDS/main/eval/eval_utils.py
code served (permissive licence) · get_code("901d2e9be65994e4")
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. 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

torch.manual_seed(0)

# Build a minimal model that satisfies the function's usage
class MinimalModel(nn.Module):
    def __init__(self, vocab_size=100, hidden_size=4):
        super().__init__()
        self.linear = nn.Linear(hidden_size, vocab_size)
        self.device_type = 'cpu'
    
    @property
    def device(self):
        class Device:
            def __init__(self, type_):
                self.type = type_
        return Device(self.device_type)
    
    def forward(self, input_ids, attention_mask):
        # Create dummy logits output
        batch_size = input_ids.shape[0]
        seq_len = input_ids.shape[1]
        logits = torch.randn(batch_size, seq_len, 100)
        
        class Output:
            pass
        
        output = Output()
        output.logits = logits
        return output

# Build a minimal tokenizer that satisfies the function's usage
class MinimalTokenizer:
    def __init__(self, vocab_size=100):
        self.vocab_size = vocab_size
        self.pad_token_id = 0
    
    def __call__(self, texts, padding='longest', return_tensors='pt', add_special_tokens=True):
        batch_size = len(texts)
        seq_len = 5
        input_ids = torch.randint(1, self.vocab_size, (batch_size, seq_len))
        attention_mask = torch.ones(batch_size, seq_len, dtype=torch.long)
        
        class TokenizerOutput:
            pass
        
        output = TokenizerOutput()
        output.input_ids = input_ids
        output.attention_mask = attention_mask
        return output
    
    def convert_ids_to_tokens(self, token_ids):
        if isinstance(token_ids, torch.Tensor):
            token_ids = token_ids.tolist()
        if isinstance(token_ids, list):
            if isinstance(token_ids[0], int):
                return [f"token_{i}" for i in token_ids]
            else:
                return [f"token_{i}" for i in token_ids]
        return f"token_{token_ids}"

model = MinimalModel(vocab_size=100, hidden_size=4)
tokenizer = MinimalTokenizer(vocab_size=100)
prompts = ["hello world", "how are you"]

ARGS = {
    'model': model,
    'tokenizer': tokenizer,
    'prompts': prompts,
    'candidate_token_ids': None,
    'batch_size': 1,
    'return_token_predictions': False,
    'add_special_tokens': True,
    'disable_tqdm': True
}
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 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: none by the harness; any seed is in the setup block
get_next_word_predictions Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
tdano1/CRDS/main/eval/utils.py
code served (permissive licence) · get_code("d134c88af0236263")
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. 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 unittest.mock import Mock

torch.manual_seed(0)

# Build a minimal model that satisfies function requirements
class MinimalLanguageModel(nn.Module):
    def __init__(self, vocab_size=100, hidden_size=4):
        super().__init__()
        self.linear = nn.Linear(hidden_size, vocab_size)
        self.device_obj = Mock()
        self.device_obj.type = 'cpu'
    
    @property
    def device(self):
        return self.device_obj
    
    def forward(self, input_ids, attention_mask):
        # input_ids: (batch_size, seq_len)
        # Return logits of shape (batch_size, seq_len, vocab_size)
        batch_size, seq_len = input_ids.shape
        hidden = torch.randn(batch_size, seq_len, 4)
        logits = self.linear(hidden)
        return Mock(logits=logits)

# Build a minimal tokenizer that satisfies function requirements
class MinimalTokenizer:
    def __init__(self, vocab_size=100):
        self.vocab_size = vocab_size
    
    def __call__(self, texts, padding='longest', return_tensors='pt', add_special_tokens=True):
        if isinstance(texts, str):
            texts = [texts]
        max_len = max(len(text.split()) for text in texts) + 2
        batch_size = len(texts)
        input_ids = torch.randint(0, self.vocab_size, (batch_size, max_len))
        attention_mask = torch.ones(batch_size, max_len)
        return Mock(input_ids=input_ids, attention_mask=attention_mask)
    
    def convert_ids_to_tokens(self, ids):
        if isinstance(ids, torch.Tensor):
            ids = ids.tolist()
        if isinstance(ids, list):
            return [f'token_{id}' for id in ids]
        return f'token_{ids}'

model = MinimalLanguageModel(vocab_size=100, hidden_size=4)
tokenizer = MinimalTokenizer(vocab_size=100)
prompts = ["Hello world", "How are you"]
candidate_token_ids = torch.tensor([5, 10, 15, 20])

ARGS = {
    'model': model,
    'tokenizer': tokenizer,
    'prompts': prompts,
    'candidate_token_ids': candidate_token_ids,
    'batch_size': 1,
    'return_token_predictions': False,
    'add_special_tokens': True,
    'disable_tqdm': True,
}
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 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: none by the harness; any seed is in the setup block
load_hf_lm_and_tokenizer Ran
ran on a synthesized input
tdano1/CRDS/main/eval/eval_utils.py
code served (permissive licence) · get_code("5128feaec154eb63")
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 float; no output shape is stated. 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 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 unittest.mock import Mock, MagicMock

torch.manual_seed(0)

# Create a minimal mock model that satisfies the function's needs
class MockModel(torch.nn.Module):
    def __init__(self):
        super().__init__()
        self.config = Mock()
        self.config.max_position_embeddings = 2048
    
    def eval(self):
        return self
    
    def half(self):
        return self
    
    def cuda(self):
        return self

# Create a minimal mock tokenizer
class MockTokenizer:
    def __init__(self):
        self.padding_side = 'left'
        self.pad_token = None
        self.eos_token = '<|endoftext|>'
        self.pad_token_id = None
        self.eos_token_id = 50256
        self.model_max_length = 1024

# Patch the transformers imports to return our mocks
import sys
from unittest.mock import patch

mock_auto_model = Mock()
mock_auto_model.from_pretrained = Mock(return_value=MockModel())

mock_auto_tokenizer = Mock()
mock_auto_tokenizer.from_pretrained = Mock(return_value=MockTokenizer())

# Patch at module import level
sys.modules['transformers'] = Mock()
sys.modules['transformers'].AutoModelForCausalLM = mock_auto_model
sys.modules['transformers'].AutoTokenizer = mock_auto_tokenizer
sys.modules['transformers'].OPTForCausalLM = Mock
sys.modules['transformers'].GPTNeoXForCausalLM = Mock

ARGS = {
    'model_name_or_path': 'gpt2',
    'tokenizer_name_or_path': None,
    'device_map': 'auto',
    'torch_dtype': 'auto',
    'load_in_8bit': False,
    'convert_to_half': False,
    'gptq_model': False,
    'use_fast_tokenizer': False,
    'padding_side': 'left'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
auto_gptq, torch, transformers — 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 float — 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
read_problems Ran
ran; our contract draft was wrong, not the code
tdano1/CRDS/main/eval/codex_humaneval/data.py
code served (permissive licence) · get_code("776a712a353130e7")
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 json
import gzip
import os
import tempfile

# Create a temporary directory for our test files
temp_dir = tempfile.mkdtemp()

# Create a minimal JSONL.GZ file with sample data
sample_data = [
    {"task_id": "HumanEval/0", "prompt": "def hello():", "canonical_solution": "return 'hello'", "test": "assert hello() == 'hello'"},
    {"task_id": "HumanEval/1", "prompt": "def add(a, b):", "canonical_solution": "return a + b", "test": "assert add(1, 2) == 3"}
]

# Write the gzipped JSONL file
test_file = os.path.join(temp_dir, "test_problems.jsonl.gz")
with gzip.open(test_file, 'wt') as f:
    for item in sample_data:
        f.write(json.dumps(item) + '\n')

# Define ARGS with the parameter for read_problems
ARGS = {
    'evalset_file': test_file
}
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 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
check_correctness Not yet run
unverified (harvested, no recorded run)
tdano1/CRDS/main/eval/codex_humaneval/execution.py
code served (permissive licence) · get_code("006afab0fbb96716")
How it ran

no invocation recorded.

encode_prompts Not yet run
unverified (harvested, no recorded run)
tdano1/CRDS/main/utils.py
code served (permissive licence) · get_code("fa98845656aa7117")
How it ran

no invocation recorded.

generate_completions Not yet run
unverified (harvested, no recorded run)
tdano1/CRDS/main/eval/eval_utils.py
code served (permissive licence) · get_code("352416cd121404a7")
How it ran

no invocation recorded.

get_domain_values Not yet run
unverified (harvested, no recorded run)
tdano1/CRDS/main/get_top_influences.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from hamishivi/automated-instruction-selection, is pointer only (licence: NONE) · get_code("a448335aa7afa0ea")
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 hamishivi/automated-instruction-selection, 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. No invocation recorded.

score_completions Not yet run
unverified (harvested, no recorded run)
tdano1/CRDS/main/eval/utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from jinzhuoran/rwku, is pointer only (licence: NONE) · get_code("d029fa02758da921")
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 jinzhuoran/rwku, 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. No invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 8 are also present in other papers whose code we hold (8 with at least one paper dated earlier, 5 the same month, 2 later) and 6 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
estimate_pass_at_k 20 other papers
18 dated earlier, 1 the same month, 1 later
2107.03374 (2021-07)
Evaluating Large Language Models Trained on Code
2207.14255 (2022-07)
Efficient Training of Language Models to Fill in the Middle
2210.14868 (2022-10)
Multi-lingual Evaluation of Code Generation Models
and 17 more not named here
get_next_word_predictions 10 other papers
9 dated earlier, 1 the same month
2401.15006 (2024-01)
Airavata: Introducing Hindi Instruction-tuned LLM
2401.16405 (2024-01)
Scaling Sparse Fine-Tuning to Large Language Models
2402.16705 (2024-02)
SelectIT: Selective Instruction Tuning for LLMs via Uncertainty-Aware Self-Reflection
and 7 more not named here
read_problems 10 other papers
7 dated earlier, 1 the same month, 2 later
2310.20329 (2023-10)
InstructCoder: Instruction Tuning Large Language Models for Code Editing
2401.08565 (2024-01)
Tuning Language Models by Proxy
2401.14196 (2024-01)
DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence
and 7 more not named here
generate_completions 6 other papers
6 dated earlier
2310.08491 (2023-10)
Prometheus: Inducing Fine-grained Evaluation Capability in Language Models
2401.10768 (2024-01)
Knowledge Verification to Nip Hallucination in the Bud
2401.15006 (2024-01)
Airavata: Introducing Hindi Instruction-tuned LLM
and 3 more not named here
score_completions 6 other papers
5 dated earlier, 1 the same month
2402.16705 (2024-02)
SelectIT: Selective Instruction Tuning for LLMs via Uncertainty-Aware Self-Reflection
2405.05008 (2024-05)
ADELIE: Aligning Large Language Models on Information Extraction
2406.10890 (2024-06)
RWKU: Benchmarking Real-World Knowledge Unlearning for Large Language Models
and 3 more not named here
generate_completions 5 other papers
5 dated earlier
2402.16705 (2024-02)
SelectIT: Selective Instruction Tuning for LLMs via Uncertainty-Aware Self-Reflection
2405.05008 (2024-05)
ADELIE: Aligning Large Language Models on Information Extraction
2410.24175 (2024-10)
Constraint Back-translation Improves Complex Instruction Following of Large Language Models
and 2 more not named here
check_correctness 3 other papers
2 dated earlier, 1 the same month
2401.16405 (2024-01)
Scaling Sparse Fine-Tuning to Large Language Models
2404.02078 (2024-04)
Advancing LLM Reasoning Generalists with Preference Trees
2602.02979 (2026-02)
CPMöbius: Iterative Coach-Player Reasoning for Data-Free Reinforcement Learning
get_domain_values 1 other paper
1 dated earlier
2503.01807 (2025-03)
Large-Scale Data Selection for Instruction Tuning

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

Data quality is a crucial factor in large language models training. While prior work has shown that models trained on smaller, high-quality datasets can outperform those trained on much larger but noisy or low-quality corpora, systematic methods for industrial-scale data selection in instruction tuning remain underexplored. In this work, we study instruction-tuning data selection through the lens of semantic representation similarity and identify a key limitation of state-of-the-art LLM encoders: they produce highly redundant semantic embeddings. To mitigate this redundancy, we propose Compressed Representation Data Selection (CRDS), a novel framework with two variants. CRDS-R applies Rademacher random projection followed by concatenation of transformer hidden-layer representations, while CRDS-W employs whitening-based dimensionality reduction to improve representational quality. Experimental results demonstrate that both variants substantially enhance data quality and consistently outperform state-of-the-art representation-based selection methods. Notably, CRDS-W achieves strong performance using only 3.5% of the data, surpassing the full-data baseline by an average of 0.71% across four datasets. Our code is available at https://github.com/tdano1/CRDS.

For agents

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

get_harvested_code_for_paper("2602.13773")
get_code_for_paper("2602.13773")
have("2602.13773")

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

curl https://syntology.ai/api/ran/2602.13773.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.13773.svg)](https://syntology.ai/paper/2602.13773)

Connect an agent — have() is free.