ATLAS: All-round Testing of Long-context Abilities across Scales
Not on the map. Why?
Abstract (arXiv)
Long-context language models now advertise context windows up to millions of tokens, yet evaluations typically report a single length or a narrow task family, masking two failure modes: performance can collapse as length grows, and strong retrieval need not transfer to downstream use. We present ATLAS, a benchmarking framework that redefines long-context evaluation as length-dependent capability profiling. ATLAS contributes three methodological principles:(i) a layered taxonomy separating foundational operations from application workloads so failures can be attributed, (ii) length-aware AUC scoring that integrates score-length curves over a fixed 8K-1M grid, replacing single-point metrics with full degradation profiles, and (iii) ATLAScore, a harmonic-mean aggregate over taxonomy categories that penalizes imbalanced profiles, with end-to-end uncertainty propagation from subset scores through the nonlinear final aggregate. We instantiate the framework across eight capability dimensions with nine auditable components and 6,438 instances, and evaluate 26 models. Gemini-3.1-Pro-Preview leads at 128K, Claude-Opus-4.6 leads at 1M. Rankings reshuffle substantially between ATLASscore@8K-128K and ATLASscore@8K-1M: 7 models move by at least two ranks, and the two taxonomy layers share only 61% of cross-model variance, with individual rank gaps up to 12 positions. These results support reporting long-context quality by capability and length, not by a single headline score.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 17 of its 30 references that are in the map. First posted on arXiv 27 May 2026.
Builds on
17 of 30 references.
References · 17
- 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
- 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
- 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
- 2023LongBench: A Bilingual, Multitask Benchmark for Long Context Understandingcited by 1,378
- 2024RULER: What's the Real Context Size of Your Long-Context Language Models?cited by 1,270
- 2020SummEval: Re-evaluating Summarization Evaluationcited by 666
- 2024LongBench v2: Towards Deeper Understanding and Reasoning on Realistic Long-context Multitaskscited by 367
- 2024∞Bench: Extending Long Context Evaluation Beyond 100K Tokenscited by 360
- 2024BABILong: Testing the Limits of LLMs with Long Context Reasoning-in-a-Haystackcited by 211
- 2023ZeroSCROLLS: A Zero-Shot Benchmark for Long Text Understandingcited by 175
- 2022SCROLLS: Standardized CompaRison Over Long Language Sequencescited by 170
- 2023Stop Uploading Test Data in Plain Text: Practical Strategies for Mitigating Data Contamination by Evaluation Benchmarkscited by 139
5 more
- 2024Can Long-Context Language Models Subsume Retrieval, RAG, SQL, and More?cited by 53
- 2025LongCodeBench: Evaluating Coding LLMs at 1M Context Windowscited by 37
- 2025Oolong: Evaluating Long Context Reasoning and Aggregation Capabilitiescited by 33
- 2024Is It Really Long Context if All You Need Is Retrieval? Towards Genuinely Difficult Long Context NLPcited by 27
- 2026AMemGym: Interactive Memory Benchmarking for Assistants in Long-Horizon Conversationscited by 20
Built on by
Citing papers per year: 2026: 2.
2 papers.