Education
Towards Estimating Bounds on the Effect of Policies under Unobserved Confounding
As many practical fields transition to provide personalized decisions, data is increasingly relevant to support the evaluation of candidate plans and policies ( e.g.,
Stepwise Alignment for Constrained Language Model Policy Optimization Akifumi Wachi Thien Q. Tran Rei Sato Takumi Tanabe Y ouhei Akimoto L Y Corporation University of Tsukuba
Safety and trustworthiness are indispensable requirements for real-world applications of AI systems using large language models (LLMs).
Distributional Preference Alignment of LLMs via Optimal Transport Igor Melnyk
Ouyang et al., 2022, Bai et al., 2022], achieves this by learning a reward model on human preference
Cooperative Hardware-Prompt Learning for Snapshot Compressive Imaging
Taking an example of hyperspectral image reconstruction, the spectral SCI [Gehm et al., 2007] can fast capture and compress 3D hyperspectral signals as
Reinforcing LLM Agents via Policy Optimization with Action Decomposition
Beginning with the simplification of flattening all actions, we theoretically explore the discrepancies between action-level optimization and this naive token-level optimization.
Saliency-driven Experience Replay for Continual Learning
We present Saliency-driven Experience Replay - SER - a biologically-plausible approach based on replicating human visual saliency to enhance classification models in continual learning settings. Inspired by neurophysiological evidence that the primary visual cortex does not contribute to object manifold untangling for categorization and that primordial saliency biases are still embedded in the modern brain, we propose to employ auxiliary saliency prediction features as a modulation signal to drive and stabilize the learning of a sequence of non-i.i.d.
A Benchmark for Evaluating Knowledge Conflicts in Large Language Models
Large language models (LLMs) have achieved impressive advancements across numerous disciplines, yet the critical issue of knowledge conflicts, a major source of hallucinations, has rarely been studied. While a few research explored the conflicts between the inherent knowledge of LLMs and the retrieved contextual knowledge, a comprehensive assessment of knowledge conflict in LLMs is still missing.
APDDv2: Aesthetics of Paintings and Drawings Dataset with Artist Labeled Scores and Comments
Building upon the initial release of APDDv1[Jin et al., 2024], our ongoing research has identified opportunities for enhancement in data scale and annotation precision.
Integrating Suboptimal Human Knowledge with Hierarchical Reinforcement Learning for Large-Scale Multiagent Systems
While agents' learning is data-driven, sampling from millions