SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.06283 · 2026

SOCKET: SOft Collision Kernel EsTimator for Sparse Attention

Anshumali Shrivastava, Agniva Chowdhury, Aditya Desai, Duy Le, Sahil Joshi, Amar Kanakamedala, Ekam Singh, Hoang Anh, Wyatt Bellinger

arXiv · PDF · Open in the Atlas

Code that ran

15 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; 2 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 12 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/2602.06283.json · badge: /api/ran/2602.06283.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
amarka8/SOCKET canonical (classifier label)
classifier basis: census
15 ran · 5 no recorded run
FunctionStatusWhere it lives
attention_mask_to_allowed_prob Ran
ran, honoured the contract we drafted
amarka8/SOCKET/kernels/socket_triton_kernels.py
pointer only (licence: NONE) · get_code("a75ad681dca07d95")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

build_results_row Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/eval/ruler_utils/scorer.py
pointer only (licence: NONE) · get_code("00ff5b090fec69bd")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

dynamically_quantize_per_channel Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/GPT-FAST/quantize.py
pointer only (licence: NONE) · get_code("55ec8fea60e1d7e7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

expand_block_index_to_heads Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
amarka8/SOCKET/kernels/block_sparse_flash_attention.py
pointer only (licence: NONE) · get_code("caaea88e7a48e0c2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

find_multiple Ran
ran, honoured the contract we drafted
amarka8/SOCKET/GPT-FAST/model.py
pointer only (licence: NONE) · get_code("aba86ffbed27ba65")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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_group_qparams Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
amarka8/SOCKET/GPT-FAST/quantize.py
pointer only (licence: NONE) · get_code("dbabfb89a4b530ac")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

logits_to_probs Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
amarka8/SOCKET/GPT-FAST/generate.py
pointer only (licence: NONE) · get_code("fd0cfe048b12c6f2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

multinomial_sample_one_no_sync Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/GPT-FAST/generate.py
pointer only (licence: NONE) · get_code("060506d7f037e907")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

pack_scales_and_zeros Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/GPT-FAST/quantize.py
pointer only (licence: NONE) · get_code("146cdcef0c195fd7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

precompute_freqs_cis Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/GPT-FAST/model.py
pointer only (licence: NONE) · get_code("26cf476e99a51b14")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

random_walk_indices Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/kernels/block_sparse_flash_attention.py
pointer only (licence: NONE) · get_code("b8b875125635b5fc")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

random_walk_indices_from_block_probs Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/kernels/block_sparse_flash_attention.py
pointer only (licence: NONE) · get_code("7ff5f8665dfef840")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

register_args_and_configs Ran
ran; our contract draft was wrong, not the code
amarka8/SOCKET/pipeline/main_utils.py
pointer only (licence: NONE) · get_code("89f735317c12c997")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

sample Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
amarka8/SOCKET/GPT-FAST/generate.py
pointer only (licence: NONE) · get_code("b0856105fbfe6f4f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

set_logger Ran
ran on a synthesized input
amarka8/SOCKET/pipeline/main_utils.py
pointer only (licence: NONE) · get_code("d86a7d51ce10c6e4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

apply_rotary_emb Not yet run
unverified (harvested, no recorded run)
amarka8/SOCKET/GPT-FAST/model.py
pointer only (licence: NONE) · get_code("1102ffb719fa5cde")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

build_sparse_list_decode Not yet run
unverified (harvested, no recorded run)
amarka8/SOCKET/GPT-FAST/kernels/sparse.py
pointer only (licence: NONE) · get_code("e5bb10b75c0b2e3c")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

get_kernel_code Not yet run
unverified (harvested, no recorded run)
amarka8/SOCKET/GPT-FAST/kernels/soft_hash_collision_loader.py
pointer only (licence: NONE) · get_code("75282665af7ff147")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

load_soft_hash_collision Not yet run
unverified (harvested, no recorded run)
amarka8/SOCKET/GPT-FAST/kernels/soft_hash_collision_loader.py
pointer only (licence: NONE) · get_code("9545ad64a8415458")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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.

sparse_attention_fwd Not yet run
unverified (harvested, no recorded run)
amarka8/SOCKET/GPT-FAST/kernels/sparse.py
pointer only (licence: NONE) · get_code("42bbf5633ac60efd")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, amarka8/SOCKET 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 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, 9 are also present in other papers whose code we hold (9 with at least one paper dated earlier, 0 the same month, 0 later) and 11 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
find_multiple 9 other papers
9 dated earlier
2302.13971 (2023-02)
LLaMA: Open and Efficient Foundation Language Models
2310.06434 (2023-10)
Whispering LLaMA: A Cross-Modal Generative Error Correction Framework for Speech Recognition
2310.08278 (2023-10)
Lag-Llama: Towards Foundation Models for Probabilistic Time Series Forecasting
and 6 more not named here
logits_to_probs 5 other papers
5 dated earlier
2402.04396 (2024-02)
QuIP#: Even Better LLM Quantization with Hadamard Incoherence and Lattice Codebooks
2406.11235 (2024-06)
QTIP: Quantization with Trellises and Incoherence Processing
2408.07092 (2024-08)
Post-Training Sparse Attention with Double Sparsity
and 2 more not named here
dynamically_quantize_per_channel 4 other papers
4 dated earlier
2308.02560 (2023-08)
From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion
2408.07092 (2024-08)
Post-Training Sparse Attention with Double Sparsity
2408.11049 (2024-08)
MagicDec: Breaking the Latency-Throughput Tradeoff for Long Context Generation with Speculative Decoding
and 1 more not named here
apply_rotary_emb 3 other papers
3 dated earlier
2408.07092 (2024-08)
Post-Training Sparse Attention with Double Sparsity
2501.06589 (2025-01)
Ladder-Residual: Parallelism-Aware Architecture for Accelerating Large Model Inference with Communication Overlapping
2505.02707 (2025-05)
Voila: Voice-Language Foundation Models for Real-Time Autonomous Interaction and Voice Role-Play
get_group_qparams 3 other papers
3 dated earlier
2308.02560 (2023-08)
From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion
2408.07092 (2024-08)
Post-Training Sparse Attention with Double Sparsity
2511.00351 (2025-11)
Reject Only Critical Tokens: Pivot-Aware Speculative Decoding
pack_scales_and_zeros 3 other papers
3 dated earlier
2308.02560 (2023-08)
From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion
2408.07092 (2024-08)
Post-Training Sparse Attention with Double Sparsity
2511.00351 (2025-11)
Reject Only Critical Tokens: Pivot-Aware Speculative Decoding
precompute_freqs_cis 1 other paper
1 dated earlier
2408.07092 (2024-08)
Post-Training Sparse Attention with Double Sparsity
sample 1 other paper
1 dated earlier
2408.07092 (2024-08)
Post-Training Sparse Attention with Double Sparsity
set_logger 1 other paper
1 dated earlier
2407.01527 (2024-07)
KV Cache Compression, But What Must We Give in Return? A Comprehensive Benchmark of Long Context Capable Approaches

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

Exploiting sparsity during long-context inference is key to scaling large language models, as attention dominates the cost of autoregressive decoding. Sparse attention reduces this cost by restricting computation to a subset of tokens, but its effectiveness depends on efficient scoring and selection at inference time. We revisit Locality-Sensitive Hashing (LSH) and introduce SOCKET, a SOft Collision Kernel EsTimator that replaces hard bucket matches with probabilistic, similarity-aware aggregation. Traditional LSH yields binary collision signals that limit ranking quality and require substantial memory to perform well. In contrast, soft LSH accumulates graded collision evidence across hash tables, preserving top-k ordering with significantly less memory. This reframes LSH from a candidate generator into a principled scoring kernel for sparse attention. Leveraging this property, SOCKET enables efficient token selection without ad hoc voting and matches or surpasses prior sparse attention methods across multiple long-context benchmarks. With a custom CUDA scoring kernel and a Flash Decode Triton backend, SOCKET achieves up to 1.5× higher throughput than FlashAttention. Code is open-sourced at https://github.com/amarka8/SOCKET.

For agents

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

get_harvested_code_for_paper("2602.06283")
get_code_for_paper("2602.06283")
have("2602.06283")

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

curl https://syntology.ai/api/ran/2602.06283.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.06283.svg)](https://syntology.ai/paper/2602.06283)

Connect an agent — have() is free.