Search Results for author:
Found 8 papers, 7 papers with code
Free Process Rewards without Process Labels
The only assumption is to parameterize the outcome reward as the log-likelihood ratios of the policy and reference models, which can be optimized regardless of the specific choice of loss objectives.
FATH: Authentication-based Test-time Defense against Indirect Prompt Injection Attacks
Unlike existing approaches that prevent LLMs from answering additional instructions in external text, our method implements an authentication system, requiring LLMs to answer all received instructions with a security policy and selectively filter out responses to user instructions as the final output.
Process Reward Model with Q-Value Rankings
PQM optimizes Q-value rankings based on a novel comparative loss function, enhancing the model's ability to capture the intricate dynamics among sequential decisions.
Position Debiasing Fine-Tuning for Causal Perception in Long-Term Dialogue
The core of the dialogue system is to generate relevant, informative, and human-like responses based on extensive dialogue history.
Reinforcement Learning with Token-level Feedback for Controllable Text Generation
To meet the requirements of real-world applications, it is essential to control generations of large language models (LLMs).
TREA: Tree-Structure Reasoning Schema for Conversational Recommendation
TREA constructs a multi-hierarchical scalable tree as the reasoning structure to clarify the causal relationships between mentioned entities, and fully utilizes historical conversations to generate more reasonable and suitable responses for recommended results.
Towards Hierarchical Policy Learning for Conversational Recommendation with Hypergraph-based Reinforcement Learning
Conversational recommendation systems (CRS) aim to timely and proactively acquire user dynamic preferred attributes through conversations for item recommendation.
DDG-DA: Data Distribution Generation for Predictable Concept Drift Adaptation
To handle concept drift, previous methods first detect when/where the concept drift happens and then adapt models to fit the distribution of the latest data.
