RuleSafe-VL: Evaluating Rule-Conditioned Decision Reasoning in Vision-Language Content Moderation

• URL: http://arxiv.org/abs/2605.07760v1
• Date: Fri, 08 May 2026 14:05:00 GMT
• Title: RuleSafe-VL: Evaluating Rule-Conditioned Decision Reasoning in Vision-Language Content Moderation
• Abstract summary: RuleSafe-VL is a benchmark for rule-conditioned decision reasoning in content moderation.<n>It formalizes 93 atomic rules and 92 typed rule relations, yielding 2,166 context-sensitive image-text cases.<n>It identifies activated rules, recover rule interactions, judge decision sufficiency, and resolve outcomes once missing context is supplied.
• Score: 4.531887731522564
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Platform content moderation applies explicit policy rules and context-dependent conditions to decide whether user content is allowed, restricted, or removed. A correct moderation outcome must therefore depend on which rules a case activates, how those rules interact, and whether the available evidence is sufficient. Current multimodal safety benchmarks largely reduce moderation to matching predefined final labels, leaving this underlying rule structure untested. As a result, a high benchmark score reveals little about whether a model applies the policy correctly or arrives at the correct label through superficial cues. To evaluate this rule-governed process, we introduce RuleSafe-VL, a benchmark for rule-conditioned decision reasoning in vision-language content moderation. Derived from publicly available platform moderation policies, RuleSafe-VL formalizes 93 atomic rules and 92 typed rule relations, yielding 2,166 context-sensitive image-text cases across three high-risk policy families. Its four diagnostic tasks decompose moderation into a rule-conditioned decision chain. They identify activated rules, recover rule interactions, judge decision sufficiency, and resolve outcomes once missing context is supplied. Experiments on 10 frontier, open-source, and safety-oriented VLMs reveal rule-relation recovery as the dominant bottleneck, where the best model reaches only 64.8 Macro-F1 and some safety-oriented models fall below 7 Macro-F1. Decision-state prediction also remains unreliable, peaking at 64.5 Macro-F1. RuleSafe-VL shifts moderation evaluation from final-label scoring toward diagnostic assessment of rule-conditioned decision reasoning.

Related papers

• Grounded Normative Rule Generation with Structured Search [47.28595196084396]
  We formalize the problem as Grounded Normative Rule Synthesis (GNRS)<n>We introduce GNRS-Search, a framework that utilizes Markov Chain Monte Carlo (MCMC) sampling to optimize a discrete, five-slot And-Or Graph (AOG)<n>We evaluate our approach on GNRS-Bench, a benchmark spanning 116 controlled goals across eight scene families, and RealCharterBench, which evaluates transfer to 53 real-derived policy tasks with hidden source clauses.
  arXiv  Detail & Related papers  (2026-08-23T05:43:42Z)
• SVR: Self-Verifying Refinement via Joint Verdict-Confidence Reinforcement Learning for Adaptive Test-Time Compute [62.3458279176813]
  Self-Verifying Refinement is an oracle-free multi-turn reinforcement learning framework.<n>It learns to use self-verification as a compute-control policy.<n>It achieves a macro-average accuracy of 0.563 with only 2.99 inference turns on average.
  arXiv  Detail & Related papers  (2026-07-30T16:20:58Z)
• SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [25.10453609307193]
  In real-world applications, guardrails are often expected to identify unsafe user-model interactions according to application-specific safety policies.<n>In this work, we study this setting under the paradigm of in-context policy guardrailing, where guardrails predict safety violations based on policy specifications provided in context.<n>We introduce SafePyramid, a safety benchmark comprising 1,000 multi-turn conversations across 10 domains and 3,000 corresponding application-specific policies, which together contain 61,699 distinct natural-language rules.
  arXiv  Detail & Related papers  (2026-06-29T07:27:22Z)
• DriveJudge: Rethinking Autonomous Driving Evaluation with Vision-Language Models [53.73898294218455]
  DriveJudge is a driving evaluation agent that combines rule-grounded evaluation with Vision-Language Model (VLM) reasoning.<n>We introduce two benchmark tasks: Driving Quality Classification and Trajectory Preference Selection.<n>DriveJudge outperformsS for driving quality classification by 21.23 AUC, and the recent VLM-based DriveCritic for trajectory preference selection by 6.5%.
  arXiv  Detail & Related papers  (2026-06-15T23:39:36Z)
• Diagnosing Live Within-Policy Instruction Conflicts in LLM Agents with Witnessed Resolution Profiles [15.65681557926802]
  LLM agents are governed by long-lived natural-collision prompt policies, but individually reasonable standing rules can interact in uninspected ways.<n>We study live intra-policy rule-conflict diagnosis: finding rule pairs inside a single prompt policy that can co-govern a realistic state.<n>We introduce WIRE, a Witnessed Intra-policy Rule Evaluation pipeline.
  arXiv  Detail & Related papers  (2026-05-27T00:09:45Z)
• Escaping the Agreement Trap: Defensibility Signals for Evaluating Rule-Governed AI [0.6138671548064355]
  We formalize evaluation as policy-grounded correctness and introduce the Defensibility Index (DI) and Ambiguity Index (AI)<n>We validate the framework on 193,000+ Reddit moderation decisions across multiple communities and evaluation cohorts.
  arXiv  Detail & Related papers  (2026-04-22T18:05:29Z)
• Do LLMs Follow Their Own Rules? A Reflexive Audit of Self-Stated Safety Policies [0.0]
  LLMs internalize safety policies through RLHF, yet these policies are never formally specified and remain difficult to inspect.<n>Existing benchmarks evaluate models against external standards but do not measure whether models understand and enforce their own boundaries.<n>We introduce the Symbolic-Neural Consistency Audit (SNCA), a framework that extracts a model's self-stated safety rules via structured prompts.
  arXiv  Detail & Related papers  (2026-04-10T10:18:45Z)
• GMP: A Benchmark for Content Moderation under Co-occurring Violations and Dynamic Rules [10.423914922203265]
  Large language models (LLMs) are adept at following fixed guidelines, but their judgment capabilities degrade when policies are unstable or context-dependent.<n>This raises a critical question for evaluation: Does high performance on existing static benchmarks truly guarantee robust generalization of AI judgment to real-world scenarios involving co-occurring violations and dynamically changing rules?
  arXiv  Detail & Related papers  (2026-03-02T10:50:11Z)
• LEC: Linear Expectation Constraints for False-Discovery Control in Selective Prediction and Routing Systems [95.35293543918762]
  Large language models (LLMs) often generate unreliable answers, while uncertainty methods fail to fully distinguish correct from incorrect predictions.<n>We address this issue through the lens of false discovery rate (FDR) control, ensuring that among all accepted predictions, the proportion of errors does not exceed a target risk level.<n>We propose LEC, which reinterprets selective prediction as a constrained decision problem by enforcing a Linear Expectation Constraint.
  arXiv  Detail & Related papers  (2025-12-01T11:27:09Z)
• Policy-Aware Generative AI for Safe, Auditable Data Access Governance [0.0]
  We present a policy aware controller that uses a large language model (LLM) to interpret natural language requests against written policies and metadata, not raw data.<n>The system, implemented with Google Gemini2.0 Flash, executes a six-stage reasoning framework with early hard policy gates and deny by default.
  arXiv  Detail & Related papers  (2025-10-27T16:10:55Z)
• Asking For It: Question-Answering for Predicting Rule Infractions in Online Content Moderation [1.803599876087764]
  ModQ is a novel question-answering framework for rule-sensitive content moderation.<n>We implement two model variants and train them on large-scale datasets from Reddit and Lemmy.<n>Both models outperform state-of-the-art baselines in identifying moderation-relevant rule violations.
  arXiv  Detail & Related papers  (2025-10-07T18:11:27Z)
• SEED-GRPO: Semantic Entropy Enhanced GRPO for Uncertainty-Aware Policy Optimization [57.69385990442078]
  Large language models (LLMs) exhibit varying levels of confidence across input prompts (questions)<n>Semantic entropy measures the diversity of meaning in multiple generated answers given a prompt and uses this to modulate the magnitude of policy updates.
  arXiv  Detail & Related papers  (2025-05-18T10:20:59Z)
• Conformal Off-Policy Evaluation in Markov Decision Processes [53.786439742572995]
  Reinforcement Learning aims at identifying and evaluating efficient control policies from data. Most methods for this learning task, referred to as Off-Policy Evaluation (OPE), do not come with accuracy and certainty guarantees. We present a novel OPE method based on Conformal Prediction that outputs an interval containing the true reward of the target policy with a prescribed level of certainty.
  arXiv  Detail & Related papers  (2023-04-05T16:45:11Z)
• Offline Policy Selection under Uncertainty [113.57441913299868]
  We consider offline policy selection as learning preferences over a set of policy prospects given a fixed experience dataset. Access to the full distribution over one's belief of the policy value enables more flexible selection algorithms under a wider range of downstream evaluation metrics. We show how BayesDICE may be used to rank policies with respect to any arbitrary downstream policy selection metric.
  arXiv  Detail & Related papers  (2020-12-12T23:09:21Z)
• Doubly Robust Off-Policy Value and Gradient Estimation for Deterministic Policies [80.42316902296832]
  We study the estimation of policy value and gradient of a deterministic policy from off-policy data when actions are continuous. In this setting, standard importance sampling and doubly robust estimators for policy value and gradient fail because the density ratio does not exist. We propose several new doubly robust estimators based on different kernelization approaches.
  arXiv  Detail & Related papers  (2020-06-06T15:52:05Z)

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.