A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Liam Barkley, Brink van der Merwe
This paper empirically evaluates prompting strategies and tool-calling agents to reduce hallucinations in LLMs, finding that simpler techniques often outperform complex ones and that tool use can increase hallucination rates.
Shuzhou Yuan, Zhan Qu, A. Kangen, et al.
This paper shows that hallucinations from LLMs, when incorporated into molecule property prediction, can significantly improve predictive accuracy for some models, challenging the view that hallucinations are purely problematic.
Unknown
SHARP is a representation-level intervention framework that modulates hallucination in LVLMs by steering internal representations.
R. Friel, Atindriyo Sanyal
Chainpoll is a high-efficacy, cost-effective method for LLM hallucination detection that outperforms existing approaches while being cheaper than GPT-4.
Unknown
This paper introduces a training framework for real-time hallucination detection in LLMs using internal states, along with the HELM benchmark for evaluation.
Unknown
This paper proposes a flexible definition of hallucination in foundation models for decision-making and reviews state-of-the-art detection and mitigation methods, including those requiring internal model access.
Unknown
HaloScope harnesses unlabeled LLM generations to improve hallucination detection accuracy by formalizing the problem and leveraging unlabeled data.
Unknown
This paper enhances uncertainty-based hallucination detection in LLMs by focusing on sentence-level detection, improving performance on accessible datasets.
Unknown
UNIHD proposes a unified framework for detecting hallucinations in multimodal large language models, addressing the fragmented landscape of existing detection methods.
Unknown
SelfCheckGPT is a zero-resource, black-box hallucination detection method for generative LLMs that uses sampling-based consistency checking.
Unknown
This paper introduces INSIDE, a generalized framework using LLMs' internal states for hallucination detection from two perspectives.
Unknown
This paper investigates hallucination detection and mitigation in language models, highlighting the importance of context and proposing an iterative retraining approach.