Vision-language models (VLMs) can estimate physical quantities such as duration, speed, and acceleration from visual observations, but existing benchmarks primarily assess overall model performance against annotated ground truth. In deployment, a key question is whether an individual prediction can be trusted when its ground truth is unavailable. Self-consistency alone may fail to capture important failure modes: a VLM may produce stable-but-wrong estimates or rely on textual priors rather than visual evidence. We formulate answer-level selective prediction for quantitative physical reasoning and propose Answer-Level Trust Selection (ATS), a post-hoc, model-agnostic framework for accepting or rejecting individual VLM predictions. ATS requires no fine-tuning, auxiliary verifier, or access to the model's internal logits. Instead, it aggregates eight interpretable behavioral diagnostic scores derived from repeated queries and controlled interventions into a unified trust score. We evaluate ATS in depth on Qwen2.5-VL-7B and across 20 VLM backbones, examining selective performance, diagnostic behavior, and targeted failure modes. Our results show that intervention-based diagnostics help identify stable-but-wrong and prior-tracking predictions that repeated agreement alone may miss. However, improved failure-case rejection can come at the cost of lower retention of correct predictions. ATS therefore complements model-level capability evaluation with answer-level reliability assessment for quantitative VLM predictions. Code will be released upon publication.
In moderation, provenance, and verification pipelines a deepfake detector's output probability is read as a degree of trust, so its calibration matters as much as raw accuracy. We reframe deepfake detection as a calibrated, self-auditing trust instrument, the Calibrated Deepfake Trust Score (CDTS), and identify what governs its trustworthiness. Our central finding is a competence-trust coupling with a sharp division: the raw score's miscalibration tracks discriminative competence almost perfectly (r = -0.98, -0.98, -0.95 across two convolutional networks and a CLIP vision transformer), and a calibrator deployed without target labels fails with competence just as tightly (r = -0.98 on the primary detector, for isotonic, Platt, and beta calibrators alike). Given target labels, by contrast, any well-specified calibrator repairs any detector, including inverted ones: in-domain calibratability is not competence-limited, and the trust failure is a distribution-shift phenomenon concentrated exactly on the low-competence generators that motivate deployment. Explanation faithfulness rises and falls on the same competence axis. We reach this conclusion after uncovering, and correcting, a tie-handling degeneracy in the standard equal-mass expected-calibration-error (ECE) estimator that fabricates strong spurious competence-calibration coupling on tie-heavy calibrated scores; the same degeneracy invalidates calibration-equity gaps we previously reported, a caution for fairness auditing. Competence is trackable without labels: batch predictive entropy flags generators with high deployed calibration error at ROC-AUC 0.99, and routing source-batches on label-free competence beats confidence-based routing precisely in the low-competence regimes the coupling identifies, while confidence regains the advantage where competence is high. Trust scoring must be competence-aware; CDTS is the mechanism.
Open platforms increasingly route tasks among heterogeneous LLM agents--differing in base model, scaffold, and tool stack--whose competence varies sharply by skill: an agent excellent at one skill may be useless at another. The standard reputation approach summarizes each agent by a single global trust score, but that scalar is the wrong object here, because routing every task to the globally most-trusted agent leaves the value of specialization unclaimed. We study skill-conditional trust R(i | k)--the trust to place in agent i for a task requiring skill k, rather than one score per agent--and pose three falsifiable questions: when is conditioning worth it, how much cross-skill evidence should be borrowed, and whether that borrowing is safe. A controlled phase-diagram analysis answers the first two: conditional trust wins only in a specific regime--high agent heterogeneity, sparse per-skill evidence, and correlated skills--and the coupling strength beta that buys this data efficiency is dual-use, because the same cross-skill borrowing is also a laundering channel. On a public benchmark of 14 genuinely heterogeneous AppWorld agents, real pools land inside the beneficial regime--a small but genuine gain, with the per-skill best agent genuinely changing across skills. We then show that an attacker with cheap evidence in one skill and none in a target skill hijacks the conditional router, driving routing regret from 0 to 0.94 on a pool our zero-cost Conditional Information Value Test (CIVT) rates GREEN--while the ungated trust verdict it contaminates reads -0.06 instead of the honest +0.19. A zero-evidence gate bounds the attack but does not eliminate it; we characterize the residual cost under an explicit budget. We do not claim Sybil-resistance--we quantify the trade-off.
Rebecca Potts, Aiden Durrant, Rick Hackney +1cs.LG
Machine learning-based predictive emissions monitoring systems offer a practical alternative to direct emissions measurement, but their deployment across gas turbine fleets is challenging when emissions labels are available for only a small subset of assets. In this work, a trust-aware probabilistic framework is proposed for fleet-level gas turbine NOx prediction under limited labelled supervision. The framework combines a multi-head recurrent prediction model with learned confidence estimation, ensemble-based uncertainty quantification, auxiliary feature prediction, feature-space distance analysis, and operating-range diagnostics. These signals are calibrated on labelled data to produce interpretable per-sample trust scores, providing indicators of prediction reliability on unlabelled turbines, supporting the identification of predictions that should be treated with greater caution during fleet-level deployment. Confidence-based filtering reduces MAE from 0.202 at full coverage to 0.070 for the highest-confidence 10\% of predictions, demonstrating that confidence estimates are meaningfully related to prediction error. Unlabelled and out-of-distribution samples exhibit increased uncertainty and reduced confidence, indicating that the framework responds appropriately to distributional shift. The results show that the proposed trust framework provides actionable reliability information for emissions prediction on unlabelled turbines, supporting more transparent and trustworthy deployment of PEMS across industrial fleets.