Negative Results

Follow this topic Ask about this topic

Momentum

13 papers in the last four weeks, up 8% on the four weeks before. 0.2% of all new papers.

Jul 6Week of Sep 21

Latest papers 127

All topics

Position: Use Sparse Autoencoders to Discover Unknowns
Jun 30, 2025Kenny Peng, Rajiv Movva, Jon Kleinberg +2Improving Sparse Autoencoders Model Discovery
1Cornell University · 2UC Berkeley
Step-by-Step Video-to-Audio Synthesis via Negative Audio Guidance
Jun 26, 2025Akio Hayakawa, Masato Ishii, Takashi Shibuya +1Audio-Video Generation Audio Understanding
Sony AI, Tokyo, Japan · Sony Group Corporation, Tokyo, Japan
Concept-as-Tree: A Controllable Synthetic Data Framework Makes Stronger Personalized VLMs
Mar 17, 2025Ruichuan An, Kai Zeng, Ming Lu +5Recent Vision-Language Models Synthetic Data
1Peking University · 2CUHK MMLab · 3Zhongguancun Academy +1
Engineering Carbon Credits Towards A Responsible FinTech Era: The Practices, Implications, and Future
Dec 22, 2024Qingwen Zeng, Hanlin Xu, Nanjun Xu +5Carbon Emissions Greenhouse Gas
School of Electrical and Computer Engineering The University of Sydney CamperdownNSW 2006 Australia. · Adelaide Business School The University of Adelaide AdelaideSA 5005 Australia. · The University of Sydney Business School The University of Sydney CamperdownNSW 2006 Australia. +1
ContrastiveCFG: Guiding Diffusion Sampling by Contrasting Positive and Negative Concepts
Nov 26, 2024Jinho Chang, Changsun Lee, Hyungjin Chung +1Classifier-Free Guidance Text-Conditioned Diffusion Model
Graduate School of Artificial Intelligence, Korea Advanced Institute of Science and Technology, Seoul, South Korea · EverEx, Seoul, South Korea
SynCo: Synthetic Hard Negatives for Contrastive Visual Representation Learning
Oct 3, 2024Nikos Giakoumoglou, Tania StathakiContrastive Learning Imagenet
Imperial College London London, UK, SW7 2AZ
Beyond Single-Negative Preference: Multi-Negative DPO for LLM-Centric Historical Entity Linking
Date pendingTien Nam Nguyen, Emanuela Boros, Ahmed Hamdi +3Backtranslation Augmented Direct Preference Optimization History
L3i, University of La Rochelle, La Rochelle, France · University of Toulouse, IRIT, Toulouse, France · University of Innsbruck, Austria +1

Jun 30, 2025, cs.LGJ/K move · Enter open · S save

Position: Use Sparse Autoencoders to Discover Unknowns

Kenny Peng, Rajiv Movva, Jon Kleinberg, Emma Pierson+1

1Cornell University · 2UC Berkeley

Improving Sparse Autoencoders Model Discovery Interpretability Negative Results Position

Open paper AskPDF arXiv

While sparse autoencoders (SAEs) have generated significant excitement, a series of negative results have added to skepticism about their usefulness. Here, we establish a conceptual distinction that reconciles competing narratives surrounding SAEs. We argue that even if SAEs may be less effective for \textit{acting on known concepts}, SAEs are especially powerful tools for \textit{discovering unknown concepts}. This distinction separates existing negative results from positive results, and suggests several classes of SAE applications. Specifically, we outline use cases for SAEs in (i) ML interpretability, explainability, fairness, auditing, and safety, and (ii) social and health sciences.

Momentum

Related topics

Latest papers 127