Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention | Paper Citation Record

Observation dc28a1ce-4597-4161-9dcb-9584a37bd18d · outbound

This paper cites GPT-4 Technical Report.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention GPT-4 Technical Report

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:51.645092Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c258e85d-d7d5-467d-b9f9-0c31bd7e49f3 · outbound

This paper cites Flamingo: a visual language model for few-shot learning.Advances in neural information processing systems, 35:23716–23736, 2022.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Flamingo: a visual language model for few-shot learning.Advances in neural information processing systems, 35:23716–23736, 2022

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:51.785416Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c82d7dd4-43d6-44ca-b411-ba80d52f59bb · outbound

This paper cites Ms marco: A human generated machine reading comprehension dataset, 2018.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Ms marco: A human generated machine reading comprehension dataset, 2018

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:52.058181Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 33a42e8a-bea0-434d-b933-e2672357d936 · outbound

This paper cites Semantic parsing on freebase from question-answer pairs.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Semantic parsing on freebase from question-answer pairs

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:52.257329Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a60ca776-5527-44d1-9f42-342092d3dbba · outbound

This paper cites Efficient Prompting Methods for Large Language Models: A Survey.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Efficient Prompting Methods for Large Language Models: A Survey

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:52.359951Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 02791ee6-c38b-436c-84c3-dc37d77283aa · outbound

This paper cites Dialogsum: A real-life scenario dialogue summarization dataset, 2021.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Dialogsum: A real-life scenario dialogue summarization dataset, 2021

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:06.148779Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 05c947ad-662b-4e17-9174-43caadec0083 · outbound

This paper cites xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Token.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Token

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:52.536544Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7a9e3740-ee1a-4d15-a2f8-5d57c6bd5380 · outbound

This paper cites Adapting language models to compress contexts.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Adapting language models to compress contexts

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:52.675674Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b08bdc3c-a77a-4ac5-8c5d-db6baba8ccdd · outbound

This paper cites Learning to compress prompt in natural language formats.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Learning to compress prompt in natural language formats

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:05.934329Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation c4b72d42-f066-475e-bb36-81ed6c3ae155 · outbound

This paper cites Selection-p: Self-supervised task-agnostic prompt compression for faithfulness and transfer- ability.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Selection-p: Self-supervised task-agnostic prompt compression for faithfulness and transfer- ability

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:05.749826Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation fc7a328e-e450-4a7a-adf9-5c62a16e14da · outbound

This paper cites Improve Student's Reasoning Generalizability through Cascading Decomposed CoTs Distillation.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Improve Student's Reasoning Generalizability through Cascading Decomposed CoTs Distillation

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:53.098230Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e5324e9f-3ae0-4e9c-ba67-c8593d095afe · outbound

This paper cites A Silver Bullet or a Compromise for Full Attention? A Comprehensive Study of Gist Token-based Context Compression.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention A Silver Bullet or a Compromise for Full Attention? A Comprehensive Study of Gist Token-based Context Compression

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:53.201220Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d59d57d1-7339-488d-aadf-792708f69320 · outbound

This paper cites Drop: A reading comprehension benchmark requiring discrete reasoning over paragraphs, 2019.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Drop: A reading comprehension benchmark requiring discrete reasoning over paragraphs, 2019

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:05.564068Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation ae270e38-818c-4833-b480-ee211a39a96b · outbound

This paper cites The Llama 3 Herd of Models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention The Llama 3 Herd of Models

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:53.471976Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 74c3761d-e0b8-42bd-86e6-abc05924cec0 · outbound

This paper cites UniICL: An Efficient Unified Framework Unifying Compression, Selection, and Generation.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention UniICL: An Efficient Unified Framework Unifying Compression, Selection, and Generation

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:53.738605Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d26a06df-723e-440b-b2a3-533c679d152f · outbound

This paper cites In-context Autoencoder for Context Compression in a Large Language Model.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention In-context Autoencoder for Context Compression in a Large Language Model

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.165282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2fc67407-ce12-41fb-92e5-9d725f3d3d58 · outbound

This paper cites In-context autoencoder for context compression in a large language model, 2023.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention In-context autoencoder for context compression in a large language model, 2023

Reference 17

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:05.383758Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 686b64b8-f947-4f45-a613-4e09da764a83 · outbound

This paper cites Samsum corpus: A human-annotated dialogue dataset for abstractive summarization.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Samsum corpus: A human-annotated dialogue dataset for abstractive summarization

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:05.243508Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 2a0980d1-ceab-47a8-8ed8-acab20480ccd · outbound

This paper cites Hipporag: Neurobiologically inspired long-term memory for large language models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Hipporag: Neurobiologically inspired long-term memory for large language models

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.405323Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 21ee3ee0-038c-4d83-b5d0-26dea4241c77 · outbound

This paper cites Constructing a multi-hop qa dataset for comprehensive evaluation of reasoning steps.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Constructing a multi-hop qa dataset for comprehensive evaluation of reasoning steps

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.452728Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 953fbe2e-2c5f-4392-b31a-24398fe3de5d · outbound

This paper cites HiAgent: Hierarchical Working Memory Management for Solving Long-Horizon Agent Tasks with Large Language Model.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention HiAgent: Hierarchical Working Memory Management for Solving Long-Horizon Agent Tasks with Large Language Model

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.490316Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7c31f94d-0c0b-475b-b5eb-d45e3e3e308a · outbound

This paper cites EXIT: Context-Aware Extractive Compression for Enhancing Retrieval-Augmented Generation.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention EXIT: Context-Aware Extractive Compression for Enhancing Retrieval-Augmented Generation

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.536782Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cfb44f13-d35c-4e04-9f05-8a31ce81edd8 · outbound

This paper cites Unsupervised dense information retrieval with contrastive learning, 2021.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Unsupervised dense information retrieval with contrastive learning, 2021

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.609171Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1fc18c11-6f6c-46ec-96af-9732117f4ce5 · outbound

This paper cites Characterizing Prompt Compression Methods for Long Context Inference.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Characterizing Prompt Compression Methods for Long Context Inference

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.679461Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 72a078ff-dce7-4276-8520-7fec798b62ca · outbound

This paper cites Mistral 7B.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Mistral 7B

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.728768Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 411b0451-b7f9-401b-8199-04253cc88ecd · outbound

This paper cites Llmlingua: Compress- ing prompts for accelerated inference of large language models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Llmlingua: Compress- ing prompts for accelerated inference of large language models

Reference 26

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:05.073521Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 667b6791-00a0-47cd-a584-3d073e4499ac · outbound

This paper cites Longllmlingua: Accelerating and enhancing llms in long context scenarios via prompt compression.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Longllmlingua: Accelerating and enhancing llms in long context scenarios via prompt compression

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:54.829428Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ace82a66-bafa-4296-ab75-b6e4c9600f62 · outbound

This paper cites FreebaseQA: A new factoid QA data set matching trivia- style question-answer pairs with Freebase.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention FreebaseQA: A new factoid QA data set matching trivia- style question-answer pairs with Freebase

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:04.934697Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 70eac42a-1498-4ac1-8fbd-93690cd1eb1c · outbound

This paper cites PubMedQA: A dataset for biomedical research question answering.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention PubMedQA: A dataset for biomedical research question answering

Reference 29

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:04.789318Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 34f82f85-e25e-41c5-a5a3-eca9c360c08b · outbound

This paper cites Triviaqa: A large scale distantly supervised challenge dataset for reading comprehension.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Triviaqa: A large scale distantly supervised challenge dataset for reading comprehension

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:55.000758Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f6fb2d67-e400-4cd9-b5ea-5384ff63da90 · outbound

This paper cites The narrativeqa reading comprehension challenge, 2017.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention The narrativeqa reading comprehension challenge, 2017

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:55.042823Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b1b0bfb2-9269-4765-93ea-67c79b312dfd · outbound

This paper cites Natural questions: A benchmark for question answering research.Transactions of the Association for Computational Linguistics, 7:452–466, 2019.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Natural questions: A benchmark for question answering research.Transactions of the Association for Computational Linguistics, 7:452–466, 2019

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:55.128212Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a682c25f-bc44-4cbe-945b-385cca6ad8ec · outbound

This paper cites Understanding and improving information preservation in prompt compression for llms.arXiv preprint arXiv:2503.19114, 2025.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Understanding and improving information preservation in prompt compression for llms.arXiv preprint arXiv:2503.19114, 2025

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:55.245580Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f1224646-8687-4c42-996c-e4cf29a8c255 · outbound

This paper cites Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:55.398680Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 75683fb7-b7e3-4acc-9472-a046705aac34 · outbound

This paper cites Compressing context to enhance inference efficiency of large language models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Compressing context to enhance inference efficiency of large language models

Reference 35

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:04.651428Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation dedcef3b-a145-4bef-8c36-514b5434f9e2 · outbound

This paper cites Prompt Compression for Large Language Models: A Survey.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Prompt Compression for Large Language Models: A Survey

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:55.750442Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 12ea812d-cae2-4c03-b3b0-40a053fb60f9 · outbound

This paper cites 500xCompressor: Generalized Prompt Compression for Large Language Models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention 500xCompressor: Generalized Prompt Compression for Large Language Models

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:55.891940Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ea08220a-bada-4625-8e57-14aa282daf3e · outbound

This paper cites Skintern: Internalizing symbolic knowledge for distilling better cot capabilities into small language models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Skintern: Internalizing symbolic knowledge for distilling better cot capabilities into small language models

Reference 38

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:04.467675Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 4a88c5df-8f8d-4f18-b8f8-805d1e010474 · outbound

This paper cites Neural-symbolic collaborative distillation: Advancing small language models for complex reasoning tasks.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Neural-symbolic collaborative distillation: Advancing small language models for complex reasoning tasks

Reference 39

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:04.235505Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 6c597c30-8bd0-4bb5-9e98-90831bcbc32c · outbound

This paper cites Awakening augmented generation: Learning to awaken internal knowledge of large language models for question answering.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Awakening augmented generation: Learning to awaken internal knowledge of large language models for question answering

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:56.340400Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c0a00e4-459d-4847-957c-cf8e437f0488 · outbound

This paper cites Let’s verify step by step.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Let’s verify step by step

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:56.443598Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 94f35f45-df62-4a84-bf7c-afaa8680eb1f · outbound

This paper cites Visual instruction tuning.Advances in neural information processing systems, 36:34892–34916, 2023.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Visual instruction tuning.Advances in neural information processing systems, 36:34892–34916, 2023

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:56.559740Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 74cd6d82-a2ae-4c13-a3a0-276ea5818223 · outbound

This paper cites A comprehensive survey on long context language modeling.arXiv preprint arXiv:2503.17407, 2025.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention A comprehensive survey on long context language modeling.arXiv preprint arXiv:2503.17407, 2025

Reference 43

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:56.752483Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 489401f2-7bb9-48ef-9948-3725b9765744 · outbound

This paper cites Lost in the middle: How language models use long contexts.Transactions of the Association for Computational Linguistics, 12:157–173, 2024.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Lost in the middle: How language models use long contexts.Transactions of the Association for Computational Linguistics, 12:157–173, 2024

Reference 44

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:56.969486Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a5e60141-3081-42f7-a13f-762ce35be0b6 · outbound

This paper cites Hybrid-Level Instruction Injection for Video Token Compression in Multi-modal Large Language Models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Hybrid-Level Instruction Injection for Video Token Compression in Multi-modal Large Language Models

Reference 45

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:57.251860Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ff4f3bea-bc49-496e-b611-c197cc88ff80 · outbound

This paper cites When not to trust language models: Investigating effectiveness of parametric and non-parametric memories.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention When not to trust language models: Investigating effectiveness of parametric and non-parametric memories

Reference 46

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:57.515359Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f6232deb-f213-4262-a23f-11eaa91a12e0 · outbound

This paper cites Learning to compress prompts with gist tokens.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Learning to compress prompts with gist tokens

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:57.767086Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3142ecc6-0433-4587-91e7-a8e8b15b0f06 · outbound

This paper cites LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression

Reference 48

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:58.091882Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ac309fc9-5957-4b9a-9859-0b87025542b4 · outbound

This paper cites PyTorch: An Imperative Style, High-Performance Deep Learning Library.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention PyTorch: An Imperative Style, High-Performance Deep Learning Library

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:58.522742Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c01779e-bf19-4c80-b353-f10c5c4e0db4 · outbound

This paper cites Know what you don’t know: Unanswerable questions for squad, 2018.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Know what you don’t know: Unanswerable questions for squad, 2018

Reference 50

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.987036Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 55abbc29-1e18-4e1c-812f-0bd8dcee7157 · outbound

This paper cites an unresolved cited work.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Unresolved cited work

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:59.221502Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0356fb1c-72f0-4f14-964e-3d578f698db3 · outbound

This paper cites Prompt programming for large language models: Beyond the few-shot paradigm.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Prompt programming for large language models: Beyond the few-shot paradigm

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:59.292733Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6d47019c-bc0f-43f3-9a49-b1d7394e7e0b · outbound

This paper cites Getting closer to ai complete question answering: A set of prerequisite real tasks.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Getting closer to ai complete question answering: A set of prerequisite real tasks

Reference 53

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.757744Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 25fbe5df-6317-4dda-826d-284efc50c4af · outbound

This paper cites Liu, and Christopher D.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Liu, and Christopher D

Reference 54

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.626251Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation f0352e15-0a93-49b9-9621-47a7bed264f7 · outbound

This paper cites TACO-RL: Task Aware Prompt Compression Optimization with Reinforcement Learning.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention TACO-RL: Task Aware Prompt Compression Optimization with Reinforcement Learning

Reference 55

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:59.528179Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 02818031-8f05-4381-a8fc-da23e2f096fe · outbound

This paper cites Large language models can be easily distracted by irrelevant context.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Large language models can be easily distracted by irrelevant context

Reference 56

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:59.647037Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c219b73b-9068-4e54-ab62-d113c86fe9fd · outbound

This paper cites The web as a knowledge-base for answering complex questions.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention The web as a knowledge-base for answering complex questions

Reference 57

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.521518Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 1fa9a041-2d6d-4aa6-b892-409774173c8c · outbound

This paper cites Commonsenseqa: A question answering challenge targeting commonsense knowledge, 2019.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Commonsenseqa: A question answering challenge targeting commonsense knowledge, 2019

Reference 58

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:59.768591Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 09050c93-b8f2-4c72-8948-3c8423df857a · outbound

This paper cites MEMORYLLM: Towards Self-Updatable Large Language Models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention MEMORYLLM: Towards Self-Updatable Large Language Models

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-07T15:15:59.967663Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 61a4012e-f1f8-4a28-a153-e9857ac50fcd · outbound

This paper cites Learning to Filter Context for Retrieval-Augmented Generation.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Learning to Filter Context for Retrieval-Augmented Generation

Reference 60

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:00.256581Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f47e92e3-5cc6-41db-84b3-fc09ebe5c629 · outbound

This paper cites an unresolved cited work.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Unresolved cited work

Reference 61

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:00.406811Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 08ffc889-8318-47ec-9dad-d0957492bf77 · outbound

This paper cites Prompt compression and contrastive conditioning for controllability and toxicity reduction in language models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Prompt compression and contrastive conditioning for controllability and toxicity reduction in language models

Reference 62

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.450887Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation c0685b70-159f-496f-9642-dfe4831835b5 · outbound

This paper cites Transformers: State-of-the-art natural language processing.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Transformers: State-of-the-art natural language processing

Reference 63

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:00.614681Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f071c63f-8d97-4d5f-8f8a-942982d08fff · outbound

This paper cites Instructing large language models to identify and ignore irrelevant conditions.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Instructing large language models to identify and ignore irrelevant conditions

Reference 64

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.386442Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation d65a9dc2-3a2f-4418-b8c8-373064d54f64 · outbound

This paper cites Unlocking efficiency in large language model inference: A comprehensive survey of speculative decoding.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Unlocking efficiency in large language model inference: A comprehensive survey of speculative decoding

Reference 65

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.302661Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation dfb3ad48-2a43-47d1-9fef-2f118a452604 · outbound

This paper cites From 128K to 4M: Efficient Training of Ultra-Long Context Large Language Models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention From 128K to 4M: Efficient Training of Ultra-Long Context Large Language Models

Reference 66

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:00.996148Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e2da6ac2-d86b-4e82-bfa6-48010969a0e5 · outbound

This paper cites Qwen2.5 Technical Report.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Qwen2.5 Technical Report

Reference 67

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:01.099375Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 992064f4-76f7-4bbf-94bc-efd275054ae4 · outbound

This paper cites WikiQA: A challenge dataset for open-domain question answering.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention WikiQA: A challenge dataset for open-domain question answering

Reference 68

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.179614Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 53c794c2-399e-4f89-a52a-8ff2dda19b65 · outbound

This paper cites Hotpotqa: A dataset for diverse, explainable multi-hop question answering.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Hotpotqa: A dataset for diverse, explainable multi-hop question answering

Reference 69

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:01.364526Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e955baa6-c129-4c3b-b9ce-f514702e393b · outbound

This paper cites Extending Llama-3's Context Ten-Fold Overnight.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Extending Llama-3's Context Ten-Fold Overnight

Reference 70

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:01.463706Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1588f1f9-d33d-4ffb-bf9e-cdff73cff3bc · outbound

This paper cites AdaComp: Extractive Context Compression with Adaptive Predictor for Retrieval-Augmented Large Language Models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention AdaComp: Extractive Context Compression with Adaptive Predictor for Retrieval-Augmented Large Language Models

Reference 71

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:01.565251Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e9b2a7ad-a09d-464c-adfa-027d2a0f65bb · outbound

This paper cites Beyond LLaVA-HD: Diving into High-Resolution Large Multimodal Models.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Beyond LLaVA-HD: Diving into High-Resolution Large Multimodal Models

Reference 72

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:01.697010Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 09d0652a-f58c-46a6-9420-6fe466ed52d7 · outbound

This paper cites LongRAG: A Dual-Perspective Retrieval-Augmented Generation Paradigm for Long-Context Question Answering.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention LongRAG: A Dual-Perspective Retrieval-Augmented Generation Paradigm for Long-Context Question Answering

Reference 73

Resolution: unresolved; no resolver link, observed 2026-08-07T15:16:01.829829Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1ce0d06f-d69c-4291-b51c-a8e66712722c · outbound

This paper cites George Claus Rankin Sir George Claus Rankin PC (12 August 1877 – 8 April 1946) was a British judge in India.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention George Claus Rankin Sir George Claus Rankin PC (12 August 1877 – 8 April 1946) was a British judge in India

Reference 74

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:03.040223Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation 2938864c-2654-43be-8299-776e647bf3e1 · outbound

This paper cites Background: [X] means the same as [D].

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Background: [X] means the same as [D]

Reference 76

Resolution: malformed identifier; raw_fallback, observed 2026-08-07T15:16:02.828984Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation bbb4a6c1-6870-4cb1-8ac3-5655ff5cda9e · outbound

This paper cites Guidelines: • The answer NA means that the paper does not involve crowdsourcing nor research with human subjects.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Guidelines: • The answer NA means that the paper does not involve crowdsourcing nor research with human subjects

Reference 77

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:02.672957Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation c6833a9b-ca83-41f6-86a0-5866a7ce7afa · outbound

This paper cites Question:.

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention Question:

Reference 1957

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:16:02.932011Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Coverage vector

Reference resolution

Outbound references

Pith citing papers