# Let's (not) just put things in Context: Test-Time Training for Long-Context LLMs

arXiv: https://arxiv.org/abs/2512.13898 · HTML: https://mlsanity.com/paper/2512.13898/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning.
- Builds on 34 of its 41 references that are in the map.
- First posted on arXiv 15 Dec 2025.

## Placed in

- LLM reasoning / [Test-time compute and efficient thinking](https://mlsanity.com/lr/test-time-compute-and-efficient-thinking/)

## Builds on (most cited first, 30 of 34)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Self-Consistency Improves Chain of Thought Reasoning in Language Models](https://mlsanity.com/paper/2203.11171/) (2022)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [Longformer: The Long-Document Transformer](https://mlsanity.com/paper/2004.05150/) (2020)
- [Qwen Technical Report](https://mlsanity.com/paper/2309.16609/) (2023)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters](https://mlsanity.com/paper/2408.03314/) (2024)
- [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context](https://mlsanity.com/paper/1901.02860/) (2019)
- [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021)
- [Big Bird: Transformers for Longer Sequences](https://mlsanity.com/paper/2007.14062/) (2020)
- [LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding](https://mlsanity.com/paper/2308.14508/) (2023)
- [RULER: What's the Real Context Size of Your Long-Context Language Models?](https://mlsanity.com/paper/2404.06654/) (2024)
- [Improving language models by retrieving from trillions of tokens](https://mlsanity.com/paper/2112.04426/) (2021)
- [Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation](https://mlsanity.com/paper/2108.12409/) (2021)
- [Atlas: Few-shot Learning with Retrieval Augmented Language Models](https://mlsanity.com/paper/2208.03299/) (2022)
- [Galactica: A Large Language Model for Science](https://mlsanity.com/paper/2211.09085/) (2022)
- [Extending Context Window of Large Language Models via Positional Interpolation](https://mlsanity.com/paper/2306.15595/) (2023)
- [YaRN: Efficient Context Window Extension of Large Language Models](https://mlsanity.com/paper/2309.00071/) (2023)
- [RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation](https://mlsanity.com/paper/2303.12570/) (2023)
- [Learning to (Learn at Test Time): RNNs with Expressive Hidden States](https://mlsanity.com/paper/2407.04620/) (2024)
- [LongRoPE: Extending LLM Context Window Beyond 2 Million Tokens](https://mlsanity.com/paper/2402.13753/) (2024)
- [Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking](https://mlsanity.com/paper/2403.09629/) (2024)
- [Scalable Best-of-N Selection for Large Language Models via Self-Certainty](https://mlsanity.com/paper/2502.18581/) (2025)
- [BookSum: A Collection of Datasets for Long-form Narrative Summarization](https://mlsanity.com/paper/2105.08209/) (2021)
- [ZeroSCROLLS: A Zero-Shot Benchmark for Long Text Understanding](https://mlsanity.com/paper/2305.14196/) (2023)
- [Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scaling](https://mlsanity.com/paper/2502.06703/) (2025)

## Built on by (most cited first, 18 of 18)

- [DeltaEvolve: Accelerating Scientific Discovery through Momentum-Driven Evolution](https://mlsanity.com/paper/2602.02919/) (2026)
- [Make Each Token Count: Towards Improving Long-Context Performance with KV Cache Eviction](https://mlsanity.com/paper/2605.09649/) (2026)
- [Gated Differentiable Working Memory for Long-Context Language Modeling](https://mlsanity.com/paper/2601.12906/) (2026)
- [The Efficiency Frontier: A Unified Framework for Cost-Performance Optimization in LLM Context Management](https://mlsanity.com/paper/2605.23071/) (2026)
- [No Time Like the Present: Agentic Test-Time Training for LLM Agents](https://mlsanity.com/paper/2607.03441/) (2026)
- [DySCO: Dynamic Attention-Scaling Decoding for Long-Context Language Models](https://mlsanity.com/paper/2602.22175/) (2026)
- [Reinforced Fast Weights with Next-Sequence Prediction](https://mlsanity.com/paper/2602.16704/) (2026)
- [Beyond Perplexity: A Behavioral Evaluation Framework for Deployment-Memory Claims in LLM Test-Time Training](https://mlsanity.com/paper/2607.00368/) (2026)
- [Self-Guided Test-Time Training for Long-Context LLMs](https://mlsanity.com/paper/2607.09415/) (2026)
- [Material Database Agent: A Multimodal Agentic Framework for Scientific Literature Mining](https://mlsanity.com/paper/2605.04278/) (2026)
- [FocuSFT: Bilevel Optimization for Dilution-Aware Long-Context Fine-Tuning](https://mlsanity.com/paper/2605.09932/) (2026)
- [Closing the Loop on Latent Reasoning via Test-Time Reconstruction](https://mlsanity.com/paper/2606.06252/) (2026)
- [Learning how to Forget: Fine-tuning for Long-Context Sparse Attention](https://mlsanity.com/paper/2608.19920/) (2026)
- [Unsupervised Post-Training of Foundation Models: A Survey](https://mlsanity.com/paper/2608.24982/) (2026)
- [CacheReforge: Bounded Recovery for Stale KV Caches under Evolving Adapters](https://mlsanity.com/paper/2609.30884/) (2026)
- [Using Context Is Not Enough: Test-Time Training for Personalized Reward Modeling](https://mlsanity.com/paper/2609.35109/) (2026)
- [Learning to Predict Distributions over Weight Updates for Test-Time Adaptation](https://mlsanity.com/paper/2610.01934/) (2026)
- [Encoding Context into LLM Weights: What Makes It Context?](https://mlsanity.com/paper/2610.17686/) (2026)

## Abstract (arXiv)

Progress on training and architecture strategies has enabled LLMs with millions of tokens in context length. However, empirical evidence suggests that such long-context LLMs can consume far more text than they can reliably use. On the other hand, it has been shown that inference-time compute can be used to scale performance of LLMs, often by generating thinking tokens, on challenging tasks involving multi-step reasoning. Through controlled experiments on sandbox long-context tasks, we find that such inference-time strategies show rapidly diminishing returns and fail at long context. We attribute these failures to score dilution, a phenomenon inherent to static self-attention. Further, we show that current inference-time strategies cannot retrieve relevant long-context signals under certain conditions. We propose a simple method that, through targeted gradient updates on the given context, provably overcomes limitations of static self-attention. We find that this shift in how inference-time compute is spent leads to consistently large performance improvements across models and long-context benchmarks. Our method leads to large 12.6 and 14.1 percentage point improvements for Qwen3-4B on average across subsets of LongBench-v2 and ZeroScrolls benchmarks. The takeaway is practical: for long context, a small amount of context-specific training is a better use of inference compute than current inference-time scaling strategies like producing more thinking tokens.