Human Feedback and Eval Paper Explorer

Match reason: Keyword overlap 3/3 across title and protocol fields.

• Jailbreak techniques for large language models (LLMs) evolve faster than benchmarks, making robustness estimates stale and difficult to compare across papers due to drift in datasets, harnesses, and judging protocols.
• We introduce JAILBREAK FOUNDRY (JBF), a system that addresses this gap via a multi-agent workflow to translate jailbreak papers into executable modules for immediate evaluation within a unified harness.

Match reason: Keyword overlap 3/3 across title and protocol fields.

• Large language model (LLM) safety evaluation is moving from content moderation to action security as modern systems gain persistent state, tool access, and autonomous control loops.
• We present AJAR, a red-teaming framework that exposes multi-turn jailbreak algorithms as callable MCP services and lets an Auditor Agent orchestrate them inside a tool-aware runtime built on Petri.

Match reason: Keyword overlap 3/3 across title and protocol fields.

• Experimental results reveal a harmful bias inherent in the standard greedy remasking strategy and identify a critical phenomenon we term Denoising-path Dependence, where the safety of early-stage tokens decisively influences the final…
• These findings also indicate that while current decoding strategies constitute a significant vulnerability, dLLMs possess a substantial intrinsic safety potential.

Match reason: Keyword overlap 3/3 across title and protocol fields.

• Token-level jailbreak attacks often produce incoherent or unreadable inputs and exhibit poor transferability, while prompt-level attacks lack scalability and rely heavily on manual effort and human ingenuity.

Match reason: Keyword overlap 3/3 across title and protocol fields.

• Large Language Model (LLM) agents increasingly use external tools for complex tasks and rely on embedding-based retrieval to select a small top-k subset for reasoning.
• We provide theoretical analysis of retrieval saturation and show on standard benchmarks that ToolFlood achieves up to a 95% attack success rate with a low injection rate (1% in ToolBench).

Match reason: Keyword overlap 3/3 across title and protocol fields.

• Key barriers include substantial cross-dialect lexical and phonological divergence, scarce synthesis-grade data, and the absence of a standardized multi-dialect evaluation benchmark.
• We further release the first standardized multi-dialect Arabic TTS benchmark, comprising over 11,000 utterances across 7 dialect subsets with manually verified transcripts.

Match reason: Keyword overlap 3/3 across title and protocol fields.

• We propose a unified training framework to improve robustness under realistic histories: (i) Teacher Error Knowledge by using Whisper large-v3 hypotheses as training-time history, (ii) Context Dropout to regularize over-reliance on history,…

Match reason: Keyword overlap 2/3 across title and protocol fields.

• Large vision-language models (VLMs) are highly vulnerable to multimodal jailbreak attacks that exploit visual-textual interactions to bypass safety guardrails.
• Rather than relying on curated safety-specific data or costly image-to-text conversion, we introduce a new formulation of the safety-relevant distributional shift induced by the visual modality.