No article text was recorded for this version.
long-context reasoning degradation
Version 1 of 1133 papers202 conceptsfrozen 2026-10-04
Papers in this version
- Blog: LIFBench: Evaluating the Instruction Following Performance and Stability of Large Language Models in Long-Context Scenarios
- Blog: LONGCODEU: Benchmarking Long-Context Language Models on Long Code Understanding
- Blog: LongCodeBench: Evaluating Coding LLMs at 1M Context Windows
- Blog: LV-Eval: A Balanced Long-Context Benchmark with 5 Length Levels Up to 256K
- Blog: Beyond a Million Tokens: Benchmarking and Enhancing Long-Term Memory in LLMs
- Blog: Context Length Alone Hurts LLM Performance Despite Perfect Retrieval
- Blog: Leave No Document Behind: Benchmarking Long-Context LLMs with Extended Multi-Doc QA
- Blog: A Real-World Benchmark for Evaluating Fine-Grained Issue Solving Capabilities of Large Language Models
- Blog: NovelQA: A Benchmark for Long-Range Novel Question Answering
- Blog: Leave No Document Behind: Benchmarking Long-Context LLMs with Extended Multi-Doc QA
- Blog: ∞Bench: Extending Long Context Evaluation Beyond 100K Tokens
- Blog: Context Length Alone Hurts LLM Performance Despite Perfect Retrieval
- Blog: Long Input Benchmark for Russian Analysis
- Blog: Michelangelo: Long Context Evaluations Beyond Haystacks via Latent Structure Queries
- Blog: SCROLLS: Standardized CompaRison Over Long Language Sequences
- Blog: Needle Threading: Can LLMs Follow Threads Through Near-Million-Scale Haystacks?
- Blog: Who Gets Cited Most? Benchmarking Long-Context Language Models on Scientific Articles
- Blog: Benchmarking Long-Context Language Models on Long Code Understanding
- Blog: NoLiMa: Long-Context Evaluation Beyond Literal Matching
- Blog: Retrieval or Global Context Understanding? On Many-Shot In-Context Learning for Long-Context Evaluation
- Blog: LongIns: A Challenging Long-context Instruction-based Exam for LLMs
- Blog: NovelQA: Benchmarking Question Answering on Documents Exceeding 200K Tokens
- Blog: Revisiting In-Context Learning with Long Context Language Models
- Blog: Same Task, More Tokens: the Impact of Input Length on the Reasoning Performance of Large Language Models
- Blog: Needle Threading: Can LLMs Follow Threads through Near-Million-Scale Haystacks?
- Blog: Can LLMs reason over extended multilingual contexts? Towards long-context evaluation beyond retrieval and haystacks
- Blog: NoLiMa: Long-Context Evaluation Beyond Literal Matching
- Blog: Eliciting In-context Retrieval and Reasoning for Long-context Large Language Models
- Blog: LogiDebrief: A Signal-Temporal Logic Based Automated Debriefing Approach with Large Language Models Integration
- Blog: RUST-BENCH: Benchmarking LLM Reasoning on Unstructured Text within Structured Tables
- Blog: The Limits of Long-Context Reasoning in Automated Bug Fixing
- Blog: BABILong: Testing the Limits of LLMs with Long Context Reasoning-in-a-Haystack
- Blog: Haystack Engineering: Context Engineering Meets the Long-Context Challenge in Large Language Models
- Blog: LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding
- Blog: NovelQA: Benchmarking Question Answering on Documents Exceeding 200K Tokens
- Blog: A Controllable Examination for Long-Context Language Models
- Blog: LoCoBench: A Benchmark for Long-Context Large Language Models in Complex Software Engineering
- Blog: When Does Divide and Conquer Work for Long Context LLM? A Noise Decomposition Framework
- Blog: FEA-Bench: A Benchmark for Evaluating Repository-Level Code Generation for Feature Implementation
- Blog: Ref-Long: Benchmarking the Long-context Referencing Capability of Long-context Language Models
- Blog: 100-LongBench: Are de facto Long-Context Benchmarks Literally Evaluating Long-Context Ability?
- Blog: Ada-LEval: Evaluating long-context LLMs with length-adaptable benchmarks
- Blog: Can LLMs Solve Longer Math Word Problems Better?
- Blog: RULER: What's the Real Context Size of Your Long-Context Language Models?
- Blog: ECon: On the Detection and Resolution of Evidence Conflicts
- Blog: LongRM: Revealing and Unlocking the Context Boundary of Reward Modeling
- Blog: TQA-Bench: Evaluating LLMs for Multi-Table Question Answering with Scalable Context and Symbolic Extension
- Blog: Reconstruction Probing
- Blog: XL2Bench: A Benchmark for Extremely Long Context Understanding with Long-range Dependencies
- Blog: Is Large Language Model Good at Triple Set Prediction? An Empirical Study
- Blog: FactGuard: Leveraging Multi-Agent Systems to Generate Answerable and Unanswerable Questions for Enhanced Long-Context LLM Extraction
- Blog: LooGLE v2: Are LLMs Ready for Real World Long Dependency Challenges?
- Blog: Towards General Continuous Memory for Vision-Language Models
- Blog: Multi-Agent Debate with Memory Masking
- Blog: xGen-small Technical Report
- Blog: Language Repository for Long Video Understanding
- Blog: DSBench: How Far Are Data Science Agents from Becoming Data Science Experts?
- Blog: LINKAGE: Listwise Ranking among Varied-Quality References for Non-Factoid QA Evaluation via LLMs
- Blog: AnaloBench: Benchmarking the Identification of Abstract and Long-context Analogies
- Blog: LongReason: A Synthetic Long-Context Reasoning Benchmark via Context Expansion
Showing 60 of 133 papers.
History & citation
Current · Permanent link
- The Effective Context Frontier of Large Language Models merged into this direction (87% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_long_context_reasoning_degradation_921cd1fa,
title = {long-context reasoning degradation},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/long-context-reasoning-degradation/rdr_921cd1fa110040178d9cf6c486ae7cd4},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}This exact version (version 1, frozen text):
@misc{lacuna_long_context_reasoning_degradation_921cd1fa_v1,
title = {long-context reasoning degradation},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/long-context-reasoning-degradation/rdr_921cd1fa110040178d9cf6c486ae7cd4/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}