What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics
- URL: http://arxiv.org/abs/2608.16370v1
- Date: Mon, 17 Aug 2026 10:21:36 GMT
- Title: What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics
- Abstract summary: We introduce a controlled runtime measurement protocol for represerving cost in a tool-using agent.<n>We vary compression severity, compare a dropping operator with a fact-acquisition operator, restore dropped state through controlled oracle interventions, and decompose tool calls into retrieval and execution.<n>Overall, compression can impose hidden interaction costs when execution-relevant state becomes absent and must be reacquired.
- Score: 0.913755431537592
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- 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. We introduce a controlled runtime measurement protocol for reacquisition cost in a bounded-horizon tool-using agent. The agent acts in a deterministic planning environment under a fixed 24-turn horizon. We vary compression severity, compare a dropping operator with a fact-preserving operator, restore dropped state through controlled oracle interventions, and decompose tool calls into retrieval and execution. We evaluate three models across two task regimes. Retrieval calls increase in all six model-regime comparisons and account for almost all added interaction; five of six remain significant after Holm correction. At the prespecified 5x comparison point, completion changes are not significant in any cell. DeepSeek shows a significant completion drop only at 10x compression. GPT-5.5 is the clearest case: completion changes from 80% to 85% (p = 1.0) while retrieval increases from 21.0 to 63.9 calls (p = .002). Retention interventions further separate state quantity, state type, and content validity. Random selection is comparable to an offline hindsight oracle, while replacing retained D-state with semantically irrelevant content increases retrieval by 57% (p < .001) without a significant completion change. In a second environment, ALFWorld, sliding compression produces no retrieval surge, showing that the reacquisition signature is environment-dependent rather than intrinsic to shortening context. Overall, compression can impose hidden interaction costs when execution-relevant state becomes absent and must be reacquired, while completion alone may not expose those costs.
Related papers
- TRACER: Per-Tool Context Retention for LLM Agents via Consequence-Attributed Reinforcement Learning [8.886237275784852]
Existing compression strategies allocate retention budgets without accounting for the downstream consequences of removing individual tool outputs.<n>We propose TRACER, which formulates compression as a sequential per-tool decision problem.
arXiv Detail & Related papers (2026-08-29T16:50:44Z) - The Sleeping Agent: What Gist-Based Context Compression Loses and Why [0.0]
We use a biologically-inspired compression framework motivated by sleep-based memory consolidation to study when compression helps and when it hurts.<n>We find gist compression substantially outperforms truncation on multi-hop reasoning and single-hop factual questions.<n>We trace this failure to a specific mechanism: the gist abstraction prompt preserves relational and event structure while discarding dates and times.
arXiv Detail & Related papers (2026-08-12T08:19:15Z) - Control Under Compression: Reliability Frontiers for Tool-Using Agents [0.0]
We introduce CompressAgent, an environment-verified benchmark for ACC compression.<n>At 75% retained context, generic rewriting and section-based compression achieve 92.7% and 92.4% success.<n>At retained-context budgets from 25% to 10%, executable protocols become fragile.
arXiv Detail & Related papers (2026-08-02T07:43:44Z) - Turning Interaction History into Execution State: A Runtime Layer for Long-Horizon Coding Agents [48.967927222079965]
Ledger is a runtime layer that distills an agent's completed interactions into an explicit execution state.<n> Ledger keeps this state in an online execution ledger and applies it at two boundaries of every step.<n>Across all 500 SWE-bench Verified instances, Ledger raises Pass@1 from 56.2% to 64.2% with GPT-5 mini and from 75.8% to 81.0% with MiniMax M2.5.
arXiv Detail & Related papers (2026-08-01T18:25:00Z) - CoACT: Action-Preserving Observation Compression for Coding Agents [8.484842324052751]
This paper proposes CoACT, an action-preserving observation compression method for coding agents.<n>CoACT is built on next-action preservation (NAP), which requires a compressed observation to induce the same next action as the raw observation.<n> Experiments on SWE-bench Verified with three agentic models show that CoACT reduces average total token consumption by 33.0%.
arXiv Detail & Related papers (2026-07-03T03:17:52Z) - A Cost-Aware, Paired Protocol for Auditing Dynamic Tool Synthesis in Agentic Video Question Answering [2.1096366377985185]
Agentic Video Questioning (VideoQA) systems invoke tools during inference, but their tool libraries are fixed.<n>We propose a cost-aware paired protocol for auditing tool-augmented video agents.
arXiv Detail & Related papers (2026-07-01T21:01:57Z) - Planning-aligned Token Compression for Long-Context Autonomous Driving [95.59023657139208]
We propose a planning-aligned working memory framework built on conditional VQ-VA.<n>Compression is conditioned on both historical trajectory and a learned planning intent.<n>We evaluate on high-signal dynamic scenarios where historical context is most critical for behavior correctness.
arXiv Detail & Related papers (2026-06-05T17:16:21Z) - Towards Retrieving Interaction Spaces for Agentic Search [62.666902610940525]
Recent direct corpus (DCI) work shows that agents can interact with the raw corpus through shell tools such as 5.4 interaction and file reads.<n>We argue that the role of retrieval for agentic search is not just to select documents that fit in the LLM context window, but to construct an interaction space.
arXiv Detail & Related papers (2026-06-05T03:47:40Z) - Scaling Laws for Agent Harnesses via Effective Feedback Compute [53.68149869349268]
emphEffective Feedback Compute (EFC) is a trace-level scaling coordinate that credits feedback only when it is informative, valid, non-redundant, and retained for subsequent decisions.<n>EFC-based coordinates consistently predict failure rates better than raw-compute baselines.
arXiv Detail & Related papers (2026-05-28T09:45:47Z) - Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses [57.20181537213498]
Agentic Harness Engineering (AHE) is a closed loop that automates harness engineering.<n>AHE addresses challenges through three matched observability pillars.<n>Ten AHE lift pass@1 on Terminal-Bench 2 from 69.7% to 77.0%, surpassing the human-designed harness Codex-CLI.<n>On SWE-bench-verified it tops aggregate success at 12% fewer tokens than the seed, and on Terminal-Bench 2 it yields +5.1 to +10.1pp cross-family gains.
arXiv Detail & Related papers (2026-04-28T16:55:02Z) - Once-for-All Sequence Compression for Self-Supervised Speech Models [62.60723685118747]
We introduce a once-for-all sequence compression framework for self-supervised speech models.
The framework is evaluated on various tasks, showing marginal degradation compared to the fixed compressing rate variants.
We also explore adaptive compressing rate learning, demonstrating the ability to select task-specific preferred frame periods without needing a grid search.
arXiv Detail & Related papers (2022-11-04T09:19:13Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.