Measuring Harness-Induced Belief Divergence in Multi-Step LLM Agents

• URL: http://arxiv.org/abs/2607.04528v1
• Date: Sun, 05 Jul 2026 22:22:39 GMT
• Title: Measuring Harness-Induced Belief Divergence in Multi-Step LLM Agents
• Abstract summary: We show that a software-agent harness can change the agent's multi-step beliefs even when the task, environment, and base LLM are fixed.<n>We introduce a belief-rollout diagnostic that elicits structured K-step trajectories over progress, risk, recoverability, constraints, failure mode, uncertainty, future success, repair cost, and next action under alternative harnesses.
• Score: 0.05371337604556311
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Software-agent benchmarks usually report whether an agent solves a task, but the agent reaches that outcome through a harness that controls what it sees, which actions it can take, which failures are repaired, which states are verified, and which evidence is logged. We show that this harness can change the agent's multi-step beliefs even when the task, environment, and base LLM are fixed. We introduce a belief-rollout diagnostic that elicits structured K-step trajectories over progress, risk, recoverability, constraints, failure mode, uncertainty, future success, repair cost, and next action under alternative harnesses. We define a cross-harness belief divergence and decompose it into an arrival term for immediate interface shifts and a growth term for horizon-dependent belief changes. On controlled coding tasks and public-benchmark stress tests, blocked actions, compressed repairs, selective verification, and cost-aware evidence pruning often preserve terminal success while changing the beliefs that drive later decisions. We further introduce BIWM, a no-training protocol that canonicalizes observations, logs censored branches, expands repair traces, records verification masks, executes risky branches in shadow, and aligns belief trajectories across harness views. The results suggest that harness design is an experimental variable in agent evaluation, not an implementation detail. Our code is available at https://github.com/Hik289/Harness-induce-bias.git.

Related papers

• Uncertainty Quantification for LLM Agents: A Taxonomy, an Evaluation Protocol, and an Empirical Study [5.818268354990671]
  Large language models act as agents that plan, call tools, retrieve evidence, maintain memory, and interact over long horizons.<n>Existing work on quantifying uncertainty for LLMs was built almost entirely for single-turn question answering.<n>This paper argues that errors and uncertainty arise from multi-turn conversations, environments, and tools rather than from a single-turn question answering setting.
  arXiv  Detail & Related papers  (2026-09-07T12:06:49Z)
• HarnessEvolve: Learning from Reference Trajectories for Reliable Agent Self-Evolution [11.431479238875712]
  We introduce HarnessEvolve, a self-evolving framework that learns from reference trajectories to achieve reliable agent self-evolution.<n>We conduct extensive experiments on several benchmarks spanning open-domain and enterprise scenarios, using different models and agent frameworks.<n>Results demonstrate that HarnessEvolve consistently outperforms state-of-the-art baselines across all benchmarks and settings.
  arXiv  Detail & Related papers  (2026-09-01T07:31:18Z)
• Beyond Suspicious Steps: Ontological Trust in Long-Horizon Agents [13.208673657449175]
  We introduce ontological trust, a task-conditioned property of trajectory prefixes, and instantiate it as RGE.<n>RGE decomposes trust along Role, Goal, and Evidence.<n>We construct a cross-domain trajectory corpus from OSWorld, FinanceBench, and EICU-AC.
  arXiv  Detail & Related papers  (2026-08-18T12:44:43Z)
• PhyAgentOS: A Self-Evolving Operating System for Embodied Agents with Decoupled Cognitive Planning and Physical Execution [49.776611937968]
  We present PhyAgentOS, a foundation delivering scheduling, verification, memory, benchmarking, and safety as system-level services.<n>Its Session-Centered treats a session, not an action, as the minimum unit of scheduling, compatibility, supervised execution, evidence collection, and acceptance.<n>SessionVerifier distinguishes execution termination from semantic task completion via evidence-grounded verdicts of success, failure, or replan.<n> Benchmarking reuses the deployment session and verification path, so results trace to real execution.
  arXiv  Detail & Related papers  (2026-07-18T04:46:53Z)
• TTHE: Test-Time Harness Evolution [50.26245555541721]
  Existing approaches optimize such harnesses before deployment, searching training or development data for a fixed agent workflow that is then frozen at test time.<n>We ask whether the harness can instead be optimized during evaluation itself, using only the unlabeled execution traces the agent produces on the test inputs.<n>During evaluation, TTHE maintains a population of candidate harnesses and refines them through an agentic proposer that reasons over their execution traces.<n>A judge then commits an improved harness from execution-derived proxy signals, and the selected program persists to govern subsequent inputs.
  arXiv  Detail & Related papers  (2026-07-09T05:53:39Z)
• ClayBuddy: A Framework, Evaluation, & Mitigation of Coding Agent Failures [1.0742675209112622]
  We study failure modes as stemming from three mechanisms: underspecification, capability errors, and agent harness errors.<n>We propose ClayBuddy, a harness that molds to user preferences and can be modified by the model in-session to mitigate these errors.
  arXiv  Detail & Related papers  (2026-06-13T18:29:53Z)
• From Failed Trajectories to Reliable LLM Agents: Diagnosing and Repairing Harness Flaws [12.757205456771295]
  This paper proposes HarnessFix, a trace-guided framework for diagnosing agent failures and repairing agent harnesses.<n> HarnessFix compiles raw execution traces and harness code into a Harness-aware Trace Intermediate Representation.<n>It attributes failures to responsible trajectory steps and harness layers, consolidates recurring diagnoses into actionable flaw records, and maps them to scoped repair operators.
  arXiv  Detail & Related papers  (2026-06-04T15:58:30Z)
• FALAT: Tracing Failures in LLM Agent Trajectories via Dependency-Guided Search [11.638321375070047]
  We propose FALAT, a diagnostic framework for failure attribution in LLM agent trajectories.<n>We evaluate FALAT on the Who&When benchmark, which includes both algorithm-generated and hand-crafted multi-agent failure trajectories.
  arXiv  Detail & Related papers  (2026-05-30T15:11:35Z)
• ExComm: Exploration-Stage Communication for Error-Resilient Agentic Test-Time Scaling [57.42714978834704]
  ExComm is a communication protocol for exploration-stage agentic test-time scaling.<n>We show that ExComm consistently outperforms strong test-time scaling baselines.
  arXiv  Detail & Related papers  (2026-05-21T07:38:44Z)
• Code as Agent Harness [107.31925305395957]
  In emerging agentic systems, code is no longer only a target output.<n>Code increasingly serves as an operational substrate for agent reasoning, acting, environment modeling, and execution-based verification.<n>This survey provides a unified roadmap toward executable, verifiable, and stateful AI agent systems.
  arXiv  Detail & Related papers  (2026-05-18T17:59:03Z)
• ContraFix: Agentic Vulnerability Repair via Differential Runtime Evidence and Skill Reuse [10.503895811137095]
  Large language model (LLM) agents are increasingly used for automated vulnerability repair.<n>Recent empirical results show that these agents still struggle with real-world vulnerabilities.<n>We present ContraFix, an agentic framework that couples runtime evidence with reusable repair skills.
  arXiv  Detail & Related papers  (2026-05-17T13:48:25Z)
• Auditing Agent Harness Safety [81.22315979618612]
  LLM agents increasingly run inside execution harnesses that dispatch tools, allocate resources, and route messages between specialized components.<n>Most safety benchmarks score only final outputs or terminal states, even though many violations occur mid-trajectory rather than at termination.<n>We propose HarnessAudit, a framework that audits full execution trajectories across boundary compliance, execution fidelity, and system stability.
  arXiv  Detail & Related papers  (2026-05-14T02:14:28Z)
• DoVer: Intervention-Driven Auto Debugging for LLM Multi-Agent Systems [48.971606069204825]
  DoVer is an intervention-driven debug framework for large language model (LLM)-based multi-agent systems.<n>It augments hypothesis generation with active verification through targeted interventions.<n>DoVer flips 18-28% of failed trials into successes, achieves up to 16% milestone progress, and validates or refutes 30-60% of failure hypotheses.
  arXiv  Detail & Related papers  (2025-12-07T09:23:48Z)
• SABER: Small Actions, Big Errors -- Safeguarding Mutating Steps in LLM Agents [52.20768003832476]
  We analyze execution traces on $$-Bench (Airline/Retail) and SWE-Bench Verified.<n>We formalize emphdecisive deviations, earliest action, level divergences that flip success to failure.<n>We introduce cm, a model-agnostic, gradient-free, test-time safeguard.
  arXiv  Detail & Related papers  (2025-11-26T01:28:22Z)

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.