Large Language Model
KALM: Knowledgeable Agent by Offline Reinforcement Learning from Large Language Model Rollouts Jing-Cheng Pang, Si-Hang Y ang, Kaiyuan Li, Xiong-Hui Chen, Nan T ang, Y ang Y u
Reinforcement learning (RL) traditionally trains agents using interaction data, which limits their capabilities to the scope of the training data.
Self-playing Adversarial Language Game Enhances LLM Reasoning
In this game, an attacker and a defender communicate around a target word only visible to the attacker.
A Natural World Text-to-Image Retrieval Benchmark
"secondary data" not reflected in their species labels [
Quantifying the Gain in Weak-to-Strong Generalization
Recent advances in large language models have shown capabilities that are extraordinary and near-superhuman.
HuRef: HUman-REadable Fingerprint for Large Language Models
However, identifying the original base model of an LLM is challenging due to potential parameter alterations. In this study, we introduce HuRef, a human-readable fingerprint for LLMs that uniquely identifies the base model without interfering with training or exposing model parameters to the public.
Gradient Cuff: Detecting Jailbreak Attacks on Large Language Models by Exploring Refusal Loss Landscapes
Large Language Models (LLMs) are becoming a prominent generative AI tool, where the user enters a query and the LLM generates an answer.
Adversarial Representation Engineering: A General Model Editing Framework for Large Language Models
Since the rapid development of Large Language Models (LLMs) has achieved remarkable success, understanding and rectifying their internal complex mechanisms has become an urgent issue. Recent research has attempted to interpret their behaviors through the lens of inner representation. However, developing practical and efficient methods for applying these representations for general and flexible model editing remains challenging.
Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms Rafael Rafailov Stanford University
A prominent issue with such methods is reward over-optimization or reward hacking, where performance as measured by the learned proxy reward model increases, but true quality plateaus or even deteriorates.