8 papers
When Less is Enough: Efficient Inference via Collaborative Reasoning
Yilei Chen, Sharut Gupta, Yannis Paschalidis +2
In this work, we introduce DUET (Dual-model Efficient Two-stage inference), a collaborative inference framework in which a capable model and a lightweight model work together to so…
Sequential-Parallel Duality in Prefix Scannable Models
Morris Yau, Sharut Gupta, Valerie Engelmayer +3
Modern neural sequence models are designed to meet the dual mandate of parallelizable training and fast sequential inference. Recent developments have given rise to various models,…
Canonicalizing Multimodal Contrastive Representation Learning
Sharut Gupta, Sanyam Kansal, Stefanie Jegelka +2
As models and data scale, independently trained networks often induce analogous notions of similarity. But, matching similarities is weaker than establishing an explicit correspond…
ReasonCACHE: Teaching LLMs To Reason Without Weight Updates
Sharut Gupta, Phillip Isola, Stefanie Jegelka +4
Can Large language models (LLMs) learn to reason without any weight update and only through in-context learning (ICL)? ICL is strikingly sample-efficient, often learning from only…
Learning Diffusion Models with Flexible Representation Guidance
Chenyu Wang, Cai Zhou, Sharut Gupta +4
Diffusion models can be improved with additional guidance towards more effective representations of input. Indeed, prior empirical work has already shown that aligning internal rep…
An Information Criterion for Controlled Disentanglement of Multimodal Data
Chenyu Wang, Sharut Gupta, Xinyi Zhang +4
Multimodal representation learning seeks to relate and decompose information inherent in multiple modalities. By disentangling modality-specific information from information that i…