# Benchmarking the Residual: What Long-Horizon Evaluations Add Beyond Matched Short-Task Performance

arXiv: https://arxiv.org/abs/2607.27283 · HTML: https://mlsanity.com/paper/2607.27283/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 37 references that are in the map.
- First posted on arXiv 29 Jul 2026.

## Builds on (most cited first, 27 of 27)

- [Let's Verify Step by Step](https://mlsanity.com/paper/2305.20050/) (2023)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments](https://mlsanity.com/paper/2404.07972/) (2024)
- [Agentless: Demystifying LLM-based Software Engineering Agents](https://mlsanity.com/paper/2407.01489/) (2024)
- [AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agents](https://mlsanity.com/paper/2407.18901/) (2024)
- [Training Software Engineering Agents and Verifiers with SWE-Gym](https://mlsanity.com/paper/2412.21139/) (2024)
- [SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?](https://mlsanity.com/paper/2509.16941/) (2025)
- [RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts](https://mlsanity.com/paper/2411.15114/) (2024)
- [Measuring AI Ability to Complete Long Software Tasks](https://mlsanity.com/paper/2503.14499/) (2025)
- [SWE-Lancer: Can Frontier LLMs Earn 1 Million from Real-World Freelance Software Engineering?](https://mlsanity.com/paper/2502.12115/) (2025)
- [NL2Repo-Bench: Towards Long-Horizon Repository Generation Evaluation of Coding Agents](https://mlsanity.com/paper/2512.12730/) (2025)
- [The Illusion of Diminishing Returns: Measuring Long Horizon Execution in LLMs](https://mlsanity.com/paper/2509.09677/) (2025)
- [SWE-EVO: Benchmarking Coding Agents in Long-Horizon Software Evolution Scenarios](https://mlsanity.com/paper/2512.18470/) (2025)
- [MASAI: Modular Architecture for Software-engineering AI Agents](https://mlsanity.com/paper/2406.11638/) (2024)
- [DeepSWE: Measuring Frontier Coding Agents on Original, Long-Horizon Engineering Tasks](https://mlsanity.com/paper/2607.07946/) (2026)
- [Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading](https://mlsanity.com/paper/2607.08964/) (2026)
- [SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks](https://mlsanity.com/paper/2603.24755/) (2026)
- [The Long-Horizon Task Mirage? Diagnosing Where and Why Agentic Systems Break](https://mlsanity.com/paper/2604.11978/) (2026)
- [SWE-Marathon: Can Agents Autonomously Complete Ultra-Long-Horizon Software Work?](https://mlsanity.com/paper/2606.07682/) (2026)
- [The Memory Curse: How Expanded Recall Erodes Cooperative Intent in LLM Agents](https://mlsanity.com/paper/2605.08060/) (2026)
- [SWE-Milestone: Evaluating AI Agents on Continuous Software Evolution](https://mlsanity.com/paper/2603.13428/) (2026)
- [SWE-Chain: Benchmarking Coding Agents on Chained Release-Level Package Upgrades](https://mlsanity.com/paper/2605.14415/) (2026)
- [On Training Large Language Models for Long-Horizon Tasks: An Empirical Study of Horizon Length](https://mlsanity.com/paper/2605.02572/) (2026)
- [ChainSWE: Benchmarking Coding Agents on Multi-Bug Software Maintenance](https://mlsanity.com/paper/2607.02606/) (2026)

## Abstract (arXiv)

Long-horizon benchmarks often show that agents fail more as tasks become longer. This observation is useful for deployment, but it does not by itself explain why failure occurs. More stages create more opportunities for ordinary errors to compound; longer tasks may also contain harder individual decisions or become harder as conversation history, tool outputs, and environment changes accumulate. We use trajectory-induced degradation to mean this last possibility: earlier execution makes later work harder. When the harmful accumulation is specifically the text visible to the model, it is often called context rot. In this position paper, we argue that to claim a "long-horizon failure", benchmarks must compare actual full-task success against a baseline prediction built from short, individual stages. We call the log-ratio between this prediction and actual success the horizon residual. The comparison must use the same agent configuration and specify in advance how stages, checkpoints, information, and budgets will be chosen. The residual shows that the full rollout differs from the chosen baseline; targeted experiments are still needed to explain why.