An approximate IPTR framework for linearly constrained optimization uses low-rank projector updates to cut per-iteration cost while preserving feasibility and convergence guarantees, with experiments showing 2.48x speedup.
Enhanc- ing LLM safety via constrained direct preference op- timization
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 1
method 1
citation-polarity summary
polarities
background 2representative citing papers
SMT achieves the highest attack success rate and HarmScore on commercial function-calling LLMs from five providers by using simulated moderation traces in multi-turn trajectories, outperforming baselines with near-minimal queries.
MGDA-Decoupled applies geometry-based multi-objective optimization within the DPO framework to find shared descent directions that account for each objective's convergence dynamics, yielding higher win rates on UltraFeedback.
Reinforced Behavior Alignment (RBA) uses self-synthesized data from a teacher LLM and reinforcement learning to close the instruction-following gap in SpeechLMs, outperforming distillation and reaching SOTA on spoken QA and speech-to-text translation benchmarks.
A survey that creates taxonomies for jailbreak attacks and defenses on LLMs, subdivides them into sub-classes, and compares evaluation approaches.
citing papers explorer
-
Scalable First-Order Interior Point Trust Region Algorithms for Linearly Constrained Optimization
An approximate IPTR framework for linearly constrained optimization uses low-rank projector updates to cut per-iteration cost while preserving feasibility and convergence guarantees, with experiments showing 2.48x speedup.
-
Beyond the Prompt: Jailbreaking Function-Calling LLMs via Simulated Moderation Traces
SMT achieves the highest attack success rate and HarmScore on commercial function-calling LLMs from five providers by using simulated moderation traces in multi-turn trajectories, outperforming baselines with near-minimal queries.
-
MGDA-Decoupled: Geometry-Aware Multi-Objective Optimisation for DPO-based LLM Alignment
MGDA-Decoupled applies geometry-based multi-objective optimization within the DPO framework to find shared descent directions that account for each objective's convergence dynamics, yielding higher win rates on UltraFeedback.
-
Enhancing Speech Large Language Models through Reinforced Behavior Alignment
Reinforced Behavior Alignment (RBA) uses self-synthesized data from a teacher LLM and reinforcement learning to close the instruction-following gap in SpeechLMs, outperforming distillation and reaching SOTA on spoken QA and speech-to-text translation benchmarks.
-
Jailbreak Attacks and Defenses Against Large Language Models: A Survey
A survey that creates taxonomies for jailbreak attacks and defenses on LLMs, subdivides them into sub-classes, and compares evaluation approaches.