9 papers
Steering by Influence: Curvature Aware Data Weighting for Activation Steering
James A. E. Dixon, Stephen J. Roberts, Francesco Quinzan
Inference-time steering offers cheap, fine-grained control over a language model's outputs by estimating a concept's representation in activation space and shifting activations tow…
Hidden in the Comments: A Context-Injection Attack Surface in Code LLMs
Noor Munir, Francesco Quinzan, Stephen Roberts
Code large language model (Code LLM) assistants generate code from heterogeneous development contexts, including open files, imported modules, pasted snippets, and comments, much o…
Detecting Contaminated Code-Generation Prompt Batches via Influence Functions
Francesco Quinzan, Noor Munir, Yishun Lu +1
Large language models (LLMs) are increasingly used for code generation, yet they remain vulnerable to prompts that elicit insecure implementations. Existing defenses typically rely…
BASIS: Batchwise Advantage Estimation from Single-Rollout Information Sharing for LLM Reasoning
Shijin Gong, Erhan Xu, Kai Ye +3
Reinforcement learning with verifiable rewards has become a standard recipe for improving the reasoning abilities of large language models. Existing algorithms face a tradeoff betw…
Learning to Orchestrate Agents under Uncertainty
Mary Chriselda Antony Oliver, Lan Jiang, Aaron Bundi Anampiu +3
Adaptive orchestration of heterogeneous agents requires making sequential delegation decisions under uncertain and evolving agent behaviour, e.g., coordinating specialised AI model…
Pruning Cannot Hurt Robustness: Certified Trade-offs in Reinforcement Learning
James Pedley, Benjamin Etheridge, Stephen J. Roberts +1
Reinforcement learning (RL) policies deployed in real-world environments must remain reliable under adversarial perturbations. At the same time, modern deep RL agents are heavily o…
Doubly Robust Alignment for Large Language Models
Erhan Xu, Kai Ye, Hongyi Zhou +3
This paper studies reinforcement learning from human feedback (RLHF) for aligning large language models with human preferences. While RLHF has demonstrated promising results, many…
Robust Reinforcement Learning from Human Feedback for Large Language Models Fine-Tuning
Kai Ye, Hongyi Zhou, Jin Zhu +2
Reinforcement learning from human feedback (RLHF) has emerged as a key technique for aligning the output of large language models (LLMs) with human preferences. To learn the reward…
When Should We Orchestrate Multiple Agents?
Umang Bhatt, Sanyam Kapoor, Mihir Upadhyay +6
Strategies for orchestrating the interactions between multiple agents, both human and artificial, can wildly overestimate performance and underestimate the cost of orchestration. W…