HarmChip is a new benchmark exposing an alignment paradox where LLMs refuse legitimate hardware security queries but comply with semantically disguised malicious requests.
hub
Direct preference optimization: Your language model is secretly a reward model
10 Pith papers cite this work. Polarity classification is still indexing.
10
Pith papers citing it
hub tools
citation-role summary
background 2
method 1
citation-polarity summary
verdicts
UNVERDICTED 10polarities
background 3representative citing papers
SRA achieves 99.71% average attack success across 26 LLMs by optimizing for coherent malicious semantics via the SRHS algorithm, with claimed theoretical guarantees on convergence and transfer.
The work creates identity-consistent synthetic makeup data via ConsistentBeauty and adapts models to real images using reinforcement learning in RealBeauty, achieving better identity preservation and real-world performance than prior methods.
PARM adapts reward models to multi-stage LLM pipelines via pipeline data and direct preference optimization, improving execution rate and solving accuracy on optimization benchmarks and showing transfer to GSM8K.
Weight Patching localizes capabilities to specific parameter modules in LLMs by replacing weights from a behavior-specialized model into a base model and validating recovery via a vector-anchor interface, revealing a hierarchy of source, routing, and execution components.
ECO uses supervised warm-up plus iterative batched DPO on a Mamba backbone to reach top neural performance on TSP and CVRP while lowering memory growth and raising throughput.
LLMs using in-context learning and fine-tuning on listener experiment data generate equalization settings that align better with population preferences than random sampling or static presets.
SCOUT uses token saliency analysis to detect both standard and contextually-plausible backdoor attacks in language models while maintaining clean accuracy.
PROGRS uses outcome-conditioned centering on PRM scores to safely integrate process rewards into GRPO for improved Pass@1 on math benchmarks.
Reflective Self-Adaptation combines failure-reflective reinforcement learning with success-guided imitation learning to enable faster and more reliable task adaptation for pre-trained Vision-Language-Action models.
citing papers explorer
-
HarmChip: Evaluating Hardware Security Centric LLM Safety via Jailbreak Benchmarking
HarmChip is a new benchmark exposing an alignment paradox where LLMs refuse legitimate hardware security queries but comply with semantically disguised malicious requests.
-
LLM-Agnostic Semantic Representation Attack
SRA achieves 99.71% average attack success across 26 LLMs by optimizing for coherent malicious semantics via the SRHS algorithm, with claimed theoretical guarantees on convergence and transfer.
-
From Synthetic to Real: Toward Identity-Consistent Makeup Transfer with Synthetic and Real Data
The work creates identity-consistent synthetic makeup data via ConsistentBeauty and adapts models to real images using reinforcement learning in RealBeauty, achieving better identity preservation and real-world performance than prior methods.
-
PARM: Pipeline-Adapted Reward Model
PARM adapts reward models to multi-stage LLM pipelines via pipeline data and direct preference optimization, improving execution rate and solving accuracy on optimization benchmarks and showing transfer to GSM8K.
-
Weight Patching: Toward Source-Level Mechanistic Localization in LLMs
Weight Patching localizes capabilities to specific parameter modules in LLMs by replacing weights from a behavior-specialized model into a base model and validating recovery via a vector-anchor interface, revealing a hierarchy of source, routing, and execution components.
-
Rethinking Efficiency in Neural Combinatorial Optimization: Batched Preference Optimization with Mamba
ECO uses supervised warm-up plus iterative batched DPO on a Mamba backbone to reach top neural performance on TSP and CVRP while lowering memory growth and raising throughput.
-
One Prompt, Many Sounds: Modeling Listener Variability in LLM-Based Equalization
LLMs using in-context learning and fine-tuning on listener experiment data generate equalization settings that align better with population preferences than random sampling or static presets.
-
SCOUT: A Defense Against Data Poisoning Attacks in Fine-Tuned Language Models
SCOUT uses token saliency analysis to detect both standard and contextually-plausible backdoor attacks in language models while maintaining clean accuracy.
-
LLM Reasoning with Process Rewards for Outcome-Guided Steps
PROGRS uses outcome-conditioned centering on PRM scores to safely integrate process rewards into GRPO for improved Pass@1 on math benchmarks.
-
Reflection-Based Task Adaptation for Self-Improving VLA
Reflective Self-Adaptation combines failure-reflective reinforcement learning with success-guided imitation learning to enable faster and more reliable task adaptation for pre-trained Vision-Language-Action models.