Medical VLMs frequently select negated options that contradict visible chest X-ray findings, achieving only ~30% accuracy on direct presence probes, but a post-hoc consistency verifier raises accuracy above 95%.
hub Canonical reference
Decodingtrust: A comprehensive assessment of trustworthiness in gpt models
Canonical reference. 71% of citing Pith papers cite this work as background.
15
Pith papers citing it
Background
71% of classified citations
hub tools
citation-role summary
background 7
citation-polarity summary
roles
background 7representative citing papers
BEAVER is the first practical deterministic verifier that maintains sound probability bounds on LLM safety properties using token tries and frontier data structures, finding 2-3x more violations than sampling at 1/10 the compute.
Fine-tuned ModernBERT-family encoders match LLM judges on F1, false negative rate, and precision-recall for harmful output detection across adversarial datasets and attack types while promising lower cost and latency.
Introduces MM-Privacy dataset and evaluations showing MLLMs leak sensitive data from images in various tasks, highlighting task inconsistency effects.
The paper introduces a paired testing protocol for batch-conditioned refusal robustness in LLM serving and reports low rates of genuine safety-label flips after adjudication, with a batch-invariant kernel ablation eliminating observed flips.
Only 39% of LLM safety benchmark repositories run without modification, 6% include ethical warnings, and adoption tracks author prominence and runnability rather than code quality metrics.
GPTFuzz is a black-box fuzzing framework that mutates seed jailbreak templates to automatically generate effective attacks, achieving over 90% success rates on models including ChatGPT and Llama-2.
Real-world jailbreak prompts collected from the wild achieve up to 0.95 attack success rates against major LLMs including GPT-4, with some persisting for over 240 days.
PCT is a reinforcement learning approach that trains LLMs for symmetric sentiment and helpfulness across paired opposing political prompts, reducing covert bias while preserving general performance.
Hybrid-DPO, which trains LLMs on preference pairs scored by a DeBERTa NLI entailment signal plus a verifier fluency score, improves NLI entailment over SFT in 11 of 15 model-domain cells and documents a persistent verbosity bias in GPT-4o-mini judging.
Training-free methods for LLM trustworthiness show inconsistent results across dimensions, with clear trade-offs in utility, robustness, and overhead depending on where they intervene during inference.
TrustLLM defines eight trustworthiness principles, creates a six-dimension benchmark, and evaluates 16 LLMs showing proprietary models generally lead but some open-source ones are close while over-calibration can hurt utility.
LLMs fail to detect hidden harmful intent, allowing systematic bypass of safety mechanisms through framing techniques, with reasoning modes often worsening the issue.
Survey of harmful fine-tuning attacks on LLMs, their variants, defense strategies, mechanical analysis, and evaluation methodologies.
The paper reviews the background, technology, applications, limitations, and future directions of OpenAI's Sora text-to-video generative model based on public information.
citing papers explorer
-
CXR-ContraBench: Benchmarking Negated-Option Attraction in Medical VLMs
Medical VLMs frequently select negated options that contradict visible chest X-ray findings, achieving only ~30% accuracy on direct presence probes, but a post-hoc consistency verifier raises accuracy above 95%.
-
BEAVER: An Efficient Deterministic LLM Verifier
BEAVER is the first practical deterministic verifier that maintains sound probability bounds on LLM safety properties using token tries and frontier data structures, finding 2-3x more violations than sampling at 1/10 the compute.
-
Do Encoders Suffice? A Systematic Comparison of Encoder and Decoder Safety Judges for LLM Adversarial Evaluation
Fine-tuned ModernBERT-family encoders match LLM judges on F1, false negative rate, and precision-recall for harmful output detection across adversarial datasets and attack types while promising lower cost and latency.
-
Unveiling Privacy Risks in Multi-modal Large Language Models: Task-specific Vulnerabilities and Mitigation Challenges
Introduces MM-Privacy dataset and evaluations showing MLLMs leak sensitive data from images in various tasks, highlighting task inconsistency effects.
-
A Paired Testing Protocol for Batch-Conditioned Refusal Robustness in LLM Serving
The paper introduces a paired testing protocol for batch-conditioned refusal robustness in LLM serving and reports low rates of genuine safety-label flips after adjudication, with a batch-invariant kernel ablation eliminating observed flips.
-
Benchmark of Benchmarks: Unpacking Influence and Code Repository Quality in LLM Safety Benchmarks
Only 39% of LLM safety benchmark repositories run without modification, 6% include ethical warnings, and adoption tracks author prominence and runnability rather than code quality metrics.
-
GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts
GPTFuzz is a black-box fuzzing framework that mutates seed jailbreak templates to automatically generate effective attacks, achieving over 90% success rates on models including ChatGPT and Llama-2.
-
"Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models
Real-world jailbreak prompts collected from the wild achieve up to 0.95 attack success rates against major LLMs including GPT-4, with some persisting for over 240 days.
-
Reducing Political Manipulation with Consistency Training
PCT is a reinforcement learning approach that trains LLMs for symmetric sentiment and helpfulness across paired opposing political prompts, reducing covert bias while preserving general performance.
-
RLearner-LLM: Balancing Logical Grounding and Fluency in Large Language Models via Hybrid Direct Preference Optimization
Hybrid-DPO, which trains LLMs on preference pairs scored by a DeBERTa NLI entailment signal plus a verifier fluency score, improves NLI entailment over SFT in 11 of 15 model-domain cells and documents a persistent verbosity bias in GPT-4o-mini judging.
-
A Systematic Study of Training-Free Methods for Trustworthy Large Language Models
Training-free methods for LLM trustworthiness show inconsistent results across dimensions, with clear trade-offs in utility, robustness, and overhead depending on where they intervene during inference.
-
TrustLLM: Trustworthiness in Large Language Models
TrustLLM defines eight trustworthiness principles, creates a six-dimension benchmark, and evaluates 16 LLMs showing proprietary models generally lead but some open-source ones are close while over-calibration can hurt utility.
-
Beyond Context: Large Language Models' Failure to Grasp Users' Intent
LLMs fail to detect hidden harmful intent, allowing systematic bypass of safety mechanisms through framing techniques, with reasoning modes often worsening the issue.
-
Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey
Survey of harmful fine-tuning attacks on LLMs, their variants, defense strategies, mechanical analysis, and evaluation methodologies.
-
Sora: A Review on Background, Technology, Limitations, and Opportunities of Large Vision Models
The paper reviews the background, technology, applications, limitations, and future directions of OpenAI's Sora text-to-video generative model based on public information.