The paper introduces a framework for diagnosing and recovering from hidden dynamics changes in deployed control policies, focusing on the problem of task readiness under dormant dynamics drift. It proposes an intervention-based Bayesian method called Evidence‑Gated Matched‑Pulse Transport that localizes faults and estimates actuator effectiveness, enabling agents to certify readiness for future tasks with limited, task‑agnostic interactions. The approach is evaluated on diverse benchmarks, measuring readiness coverage, selective risk, interaction cost, and return, and identifies regimes where transported evidence is decisive.
arXiv:2609.20973v1 Announce Type: cross Abstract: Complex problem solving (CPS) with large language models (LLMs) is often framed as a matter of stronger reasoning or longer generation. Yet early-ste...
arXiv:2605. 22949v3 Announce Type: replace Abstract: Foundation-model pools are increasingly used as black-box responders in coordinated systems where a coordinator must decide which response to trust.
arXiv:2607. 17188v2 Announce Type: replace Abstract: While test-time scaling improves the problem-solving ability of large reasoning models (LRMs) through additional inference-time computation, it can also exacerbate overthinking and underthinking, which we formulate as reasoning state--action mismatch.
arXiv:2609.13543v1 Announce Type: new Abstract: LLM agents are predominantly benchmarked on short, single-task trajectories, yet real deployments run for hours under contention, surfacing a different...
The paper introduces a diagnostic framework for long‑horizon security LLM agents that uses checkpoints to distinguish failures occurring before and after a model’s capability is exposed, and applies controlled interventions to pinpoint upstream bottlenecks. The methodology is tested on four task families—delayed reuse of discovered information, reuse of observed state, recovery from failed strategies, and decision making after uncertain outcomes—revealing that many failures happen before the agent observes the state it later needs to reuse. Experiments with Gemini 2.5 Flash and Gemini 3.7 Flash show that targeted protocol‑disambiguation guidance can significantly alter state observation rates and that the primary source of failure can shift across model generations, underscoring the need for fine‑grained failure diagnostics rather than relying solely on overall task success.
ObserverBench is a benchmark framework that evaluates whether internal mechanistic estimators—called observers—are suitable for guiding interventions, control, or safety actions in language models. It separates estimation accuracy from the loss incurred by the chosen action, showing that accurate predictions do not always lead to better decisions. Experiments on GPT‑2‑small, Qwen2.5‑7B, Gemma‑2‑9B‑it, and Qwen3.5‑9B demonstrate that observers trained on action loss can reduce deployment loss, while traditional metrics like AUROC may rank monitors differently from actual performance.
arXiv:2606. 29654v1 Announce Type: new Abstract: Multi-agent deliberation among LLMs can improve reasoning, but deployment requires deciding when the current answer is reliable enough to act on and when it should be escalated to human review.
arXiv:2608. 15101v1 Announce Type: new Abstract: Policy evaluation often estimates direct benefits and costs while treating the institutional environment as fixed.
ObserverBench is a benchmark framework that evaluates whether internal mechanistic estimators—called observers—are suitable for guiding interventions, control, or safety actions in language models. It separates estimation accuracy from the loss incurred by the chosen action, demonstrating that accurate average estimates can still lead to poor decisions. Experiments on GPT‑2‑small, Qwen2.5‑7B, Gemma‑2‑9B‑it, and Qwen3.5‑9B show that observers trained on action loss tend to select lower‑loss actions, while traditional metrics like AUROC can rank monitors differently from deployment loss, highlighting the need for task‑specific evaluation.
CARE‑VI introduces a framework for improving value targets in off‑policy actor‑critic learning by combining Conservative Adaptive Ranking and Screening (CARS), Selector‑Evaluator Value Assessment (SEVA), and Dynamic Adaptive Risk‑aware Enhancement (DARE). CARS limits candidate actions to a budgeted prefix and expands it only when uncertainty exceeds a threshold; SEVA orders candidates with selector critics and reviews their values with an evaluator critic, capping the value at the selector reference; DARE adjusts residual corrections based on candidate reliability and signal gaps. Theoretical analysis bounds errors in each component, and empirical tests on SAC, TD3, and TD7 across four MuJoCo tasks show CARE‑VI consistently outperforms baselines in mean return.
arXiv:2606. 28939v1 Announce Type: new Abstract: Behavior-cloned diffusion policies are expressive but remain vulnerable to covariate shift: small deviations from demonstrated states can compound into task failure.
arXiv:2608. 03222v1 Announce Type: cross Abstract: Software engineering (SWE) agents resolve repository-level issues through long trajectories that grow increasingly expensive as context accumulates.