Tool-using language-model agents are governed not only by task prompts but also by persistent system-side instructions that specify tools, arguments, policies, execution protocols, and recovery. Compressing these agent control contexts (ACCs) can reduce input cost and context use, yet existing prompt-compression evaluations do not reveal whether the resulting control remains operationally reliable.
arXiv:2609.36526v1 Announce Type: cross Abstract: Long-horizon agents require context compression to manage growing interaction histories. Compression quality, however, is ultimately determined by do...
The paper introduces Dynamic Tool Output Compression (DTOC), a framework that manages context in large‑language‑model agents by storing full tool outputs in external memory and inserting compact placeholders into the active context. DTOC treats context updates as explicit, reversible operations within the agent’s reasoning loop, allowing selective reconstruction of compressed outputs when needed. Experiments on the DeepSWE benchmark show that for responsive models such as Sonnet 4.6 and GPT‑5.4, DTOC reduces input tokens and agent steps while significantly improving solve rates and lowering cost per solved task, with ablation studies confirming the importance of reversibility for maintaining performance.
arXiv:2609.06059v1 Announce Type: new Abstract: As large language models evolve from question-answering systems into general-purpose agents, evaluation must move beyond static answer correctness to a...
arXiv:2608. 06503v1 Announce Type: new Abstract: Recurrent context compression controls context growth in long-horizon agents, but its behavioral effects remain poorly understood.
UniACE is a unified framework that standardizes the evaluation of large language model (LLM) agents by representing each benchmark as an instruction–tool–environment triplet and running models through a shared, task‑agnostic harness in isolated runtimes. It preserves native success criteria, offers an offline mode for dynamic‑resource tasks, and standardizes efficiency metrics, execution records, and failure attribution. Applying UniACE to 7 benchmarks across 24 domains and 15 models revealed significant score shifts, ranking reversals, and sensitivity to evidence representation, highlighting the impact of evaluation configuration on reported agent performance.
The study investigates how individual components of a coding harness—planning, action space, and context management—affect autonomous coding agents’ performance. By fixing the execution loop and varying these components across 176 settings on SWE‑Bench Verified and Terminal‑Bench 2.1, the authors find that context management is most valuable when context windows are tight, staging rule‑based elision before LLM summarization yields the best efficiency, planning serves as an accuracy scaffold for weaker models and a cost saver for stronger ones, and predefined tools help models with limited bash skills while bash‑capable models benefit from a bash‑only interface. Trajectory‑level analysis shows that context management lengthens execution paths, planning alters where trajectories terminate, and the action space determines code granularity, offering a modular framework for future harness design.
The paper evaluates five context‑trimming strategies for agentic large language model workflows, comparing them on metrics such as task success, protocol adherence, token savings, and latency. Conventional trimming methods save about 60% of tokens but achieve lower success rates, while protocol‑aware trimming raises success to 92.2% and adaptive guardrails further improve it to 96% success with 56% token savings. The study shows that preserving protocol‑critical state is more important than aggressive token removal, and that adaptive guardrails enhance efficiency, scalability, and reliability for long‑horizon agentic systems.
arXiv:2606. 02965v2 Announce Type: replace Abstract: As large language models gain tool access and are deployed as autonomous agents capable of editing records, executing transactions, and modifying infrastructure, we still evaluate them based on the sole metric of task completion.
arXiv:2608. 05604v1 Announce Type: cross Abstract: Large Language Models (LLMs) increasingly act as agents whose procedural knowledge is stored in reusable skill packages and loaded at inference time.
arXiv:2608. 13867v1 Announce Type: cross Abstract: AI coding agents are commonly evaluated as models but deployed as systems.
RefactorPlatform is an open‑source harness that standardizes the evaluation of repository‑scale refactoring agents by fixing the environment and systematically varying design choices such as model backbone, execution regime, and prompt specificity. Each run operates in an isolated workspace, logs detailed telemetry, and verifies changes with AST‑based checks. Experiments on 100 RefactorBench tasks show that AST‑aware chunking improves performance by 25‑30%, a lean retrieval‑augmented single agent outperforms a sub‑agent configuration, and retrieval’s accuracy gains offset its token overhead, keeping cost per successful refactoring unchanged.
arXiv:2608. 16370v1 Announce Type: new Abstract: Task completion is the standard metric for evaluating context compression, yet it is incomplete: compression can increase an agent's interaction cost by forcing it to reacquire dropped state while leaving completion statistically unchanged.