Paper Citation Record · LEDGER

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding

As of 9 August 2026, this Paper Citation Record lists 39 of 39 outbound references and 1 inbound Pith citation observation for arXiv:2502.08363.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2502.08363 v3

Coverage vector

measured 39 of 39 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-08T05:29:16.784078Z

measured 40 of 40 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-09T06:31:02.800959+00:00

measured 1 of 1 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-07T10:24:27.792499Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: pith, observed 2026-08-07T10:24:28.112713Z

Reference resolution

39 of 39 outbound references displayed

verified exact2
verified fuzzy13
unresolved23
parse uncertain0
malformed identifier0
metadata mismatch1

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 3ce9d62a-20b7-4a70-b208-94c4bba7ea07 · outbound

This paper cites Attention is all you need.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Attention is all you need

Reference 1

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.789630Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 19137237-9b8f-4934-9001-9d42ebcc97e4 · outbound

This paper cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.610936Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 98c36a40-4c8c-4f14-898e-9ce732a67f41 · outbound

This paper cites On the com- putational complexity of self-attention.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding On the com- putational complexity of self-attention

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.774304Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b62c9e13-2e56-4fd5-a6bb-8626fc00aa5a · outbound

This paper cites Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.621528Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 637844f6-9885-48c7-8893-86eeffabef29 · outbound

This paper cites Beyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Models.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Beyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Models

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.626580Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 601a47ac-df43-41d3-9936-3a96800a11a5 · outbound

This paper cites A survey on sparsity exploration in transformer- based accelerators.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding A survey on sparsity exploration in transformer- based accelerators

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.758528Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 7071b526-fca2-4e43-b591-d3875491b258 · outbound

This paper cites Memory-efficient Transformers via Top-$k$ Attention.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Memory-efficient Transformers via Top-$k$ Attention

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.636662Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dabb081f-ce5f-4125-99cb-800643504c8b · outbound

This paper cites What Language Model Architecture and Pretraining Objective Work Best for Zero-Shot Generalization?.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding What Language Model Architecture and Pretraining Objective Work Best for Zero-Shot Generalization?

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.641264Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ba39910-acde-41cc-a940-f8fb3a597d4e · outbound

This paper cites Roformer: Enhanced transformer with rotary position embedding.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Roformer: Enhanced transformer with rotary position embedding

Reference 9

Resolution: verified exact; arxiv_id_nonexistent, observed 2026-08-08T05:29:17.478688Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 615264ff-5416-4099-a9c5-46e13a6d8cfd · outbound

This paper cites Analyzing the Structure of Attention in a Transformer Language Model.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Analyzing the Structure of Attention in a Transformer Language Model

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.650454Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 728b6bad-1804-42b4-8bc9-ac2a2d11eee7 · outbound

This paper cites Linear Log-Normal Attention with Unbiased Concentration.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Linear Log-Normal Attention with Unbiased Concentration

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.655553Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b92d5962-84f4-4bbe-a72c-02ea9c1698b1 · outbound

This paper cites Keep the Cost Down: A Review on Methods to Optimize LLM' s KV-Cache Consumption.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Keep the Cost Down: A Review on Methods to Optimize LLM' s KV-Cache Consumption

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.660204Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 639b9220-0392-43b8-9a64-11d00946dc34 · outbound

This paper cites GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.665033Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6c6494eb-a38d-4549-9439-d9fea5712613 · outbound

This paper cites Parallel Top-K algorithms on GPU: A comprehensive study and new methods.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Parallel Top-K algorithms on GPU: A comprehensive study and new methods

Reference 14

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.743078Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation aa2206f7-abf5-495f-a5f3-94092c85929c · outbound

This paper cites Towards Efficient Generative Large Language Model Serving: A Survey from Algorithms to Systems.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Towards Efficient Generative Large Language Model Serving: A Survey from Algorithms to Systems

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.679116Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a09db41a-2dcb-4fa4-b7d8-1e02dd5cdab8 · outbound

This paper cites FlashAttention: Fast and memory-efficient exact attention with io-awareness.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding FlashAttention: Fast and memory-efficient exact attention with io-awareness

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.727900Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b26ea330-fbae-4c81-b355-2e528e481900 · outbound

This paper cites A framework for few-shot language model evaluation, 07 2024.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding A framework for few-shot language model evaluation, 07 2024

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.688092Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ca32d981-743b-460b-a7a4-0186541da4ee · outbound

This paper cites Evaluating Large Language Models Trained on Code.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Evaluating Large Language Models Trained on Code

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.692548Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9107e1b7-5994-47dd-876b-5d94b8cfb40b · outbound

This paper cites LongBench: A bilingual, multitask benchmark for long context understanding.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding LongBench: A bilingual, multitask benchmark for long context understanding

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.696976Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 26fa3b79-a1e5-404a-8f79-51db2044e5b5 · outbound

This paper cites Llama 2: Open Foundation and Fine-Tuned Chat Models.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Llama 2: Open Foundation and Fine-Tuned Chat Models

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.701409Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9fcd8ac7-c9ec-48d4-a579-585da0a813c6 · outbound

This paper cites The Llama 3 Herd of Models.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding The Llama 3 Herd of Models

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.706212Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 264105b2-83a2-467e-be50-61ae338de677 · outbound

This paper cites Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.710874Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ee629974-95d3-4019-977e-6b6c8c24f1ed · outbound

This paper cites Quantization variation: A new perspective on training transformers with low-bit precision.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Quantization variation: A new perspective on training transformers with low-bit precision

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.703856Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 1c365f05-9bc0-496c-9586-1a2cb3541729 · outbound

This paper cites Energon: Toward efficient acceleration of transformers using dynamic sparse attention.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Energon: Toward efficient acceleration of transformers using dynamic sparse attention

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.686058Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 367c63e0-033b-49dc-916d-c7ac9784b4bc · outbound

This paper cites SpAtten: Efficient sparse attention architecture with cascade token and head pruning.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding SpAtten: Efficient sparse attention architecture with cascade token and head pruning

Reference 25

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.670534Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 91512def-a04d-4598-85eb-60786a70ca1c · outbound

This paper cites A$^3$: Accelerating Attention Mechanisms in Neural Networks with Approximation.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding A$^3$: Accelerating Attention Mechanisms in Neural Networks with Approximation

Reference 26

Resolution: metadata mismatch; local_arxiv, observed 2026-08-08T05:29:16.912922Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 9892386f-9661-4b1d-80ec-97f60e102f1a · outbound

This paper cites SparQ Attention: Bandwidth-Efficient LLM Inference.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding SparQ Attention: Bandwidth-Efficient LLM Inference

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.733334Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 212781a5-ae7d-45fd-8a50-adc79ec400e0 · outbound

This paper cites Swin transformer: Hierarchical vision transformer using shifted windows.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Swin transformer: Hierarchical vision transformer using shifted windows

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.738176Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 01f7a4d1-21bc-4eda-83aa-084d997c6cb6 · outbound

This paper cites Learned token pruning for transformers.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Learned token pruning for transformers

Reference 29

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.645070Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation cb67c1a9-6f11-4b12-a3a9-65ab31363235 · outbound

This paper cites Sparser is faster and less is more: Efficient sparse attention for long-range transformers, 2024.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Sparser is faster and less is more: Efficient sparse attention for long-range transformers, 2024

Reference 30

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.628531Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation da3bffc8-1db2-4d0b-85d0-ec7434f2e2eb · outbound

This paper cites From softmax to sparsemax: A sparse model of attention and multi-label classification.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding From softmax to sparsemax: A sparse model of attention and multi-label classification

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.613503Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 1b3555f9-10c0-4b32-90e3-ac2169f76169 · outbound

This paper cites Sparse Sequence-to-Sequence Models.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Sparse Sequence-to-Sequence Models

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.755455Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e453c126-8e4f-44f8-8a04-502487eb7702 · outbound

This paper cites Sparse Attention with Linear Units.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Sparse Attention with Linear Units

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.760019Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 41b0355e-a9fd-443e-ae94-37c77c7b91df · outbound

This paper cites Longformer: The Long-Document Transformer.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Longformer: The Long-Document Transformer

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.765072Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 42c88afe-2acb-4fed-bbfd-833d130dcbb1 · outbound

This paper cites Length-Adaptive Transformer: Train Once with Length Drop, Use Anytime with Search.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Length-Adaptive Transformer: Train Once with Length Drop, Use Anytime with Search

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-08T05:29:16.769723Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8851141e-3791-4767-bf7a-2031d042e8f8 · outbound

This paper cites They are conditionally independent given the input X from which they were originally computed via V = XW V.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding They are conditionally independent given the input X from which they were originally computed via V = XW V

Reference 37

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.598968Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 037690e9-96df-49fb-83e6-f4c63908990c · outbound

This paper cites an unresolved cited work.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding Unresolved cited work

Reference 38

Resolution: unresolved; raw_fallback, observed 2026-08-08T05:29:17.583556Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation f81260cb-9103-4e1a-b13d-90a6c94cb1dc · outbound

This paper cites F Evaluation statistics In this section, we present again the experimental results from Section 4.1; however, to demonstrate statistical significance, we show the error bars.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding F Evaluation statistics In this section, we present again the experimental results from Section 4.1; however, to demonstrate statistical significance, we show the error bars

Reference 39

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T05:29:17.567937Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 4496d8bb-58b6-4e44-98d1-55bc911016f7 · outbound

This paper cites ISBN 9798400701092.

Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding ISBN 9798400701092

Reference 2023

Resolution: verified exact; arxiv_id_nonexistent, observed 2026-08-08T05:29:17.231661Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Pith citing papers

Observation ebca9c3e-09b4-4cd2-b792-5828cdc7c8a8 · inbound

Power Law Guided Dynamic Sifting for Efficient Attention cites this paper.

Power Law Guided Dynamic Sifting for Efficient Attention Top-Theta Attention: Sparsifying Transformers by Compensated Thresholding

Reference 3

Resolution: verified exact; local_arxiv, observed 2026-08-07T10:24:28.119897Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.