5 citations · 9 across the 12 of their papers we have counts for
13 papers
Scalable Circuit Learning for Interpreting Large Language Models
Naiyu Yin, Dennis Wei, Tian Gao +3
A prominent research direction in mechanistic interpretability is learning sparse circuits over LLM components to reveal how they jointly produce model behavior. However, raw neuro…
CoFrGeNet: Continued Fraction Architectures for Language Generation
Amit Dhurandhar, Vijil Chenthamarakshan, Dennis Wei +3
Transformers are arguably the preferred architecture for language generation. In this paper, inspired by continued fractions, we introduce a new function class for generative model…
Agentic AI Needs a Systems Theory
Erik Miehling, Karthikeyan Natesan Ramamurthy, Kush R. Varshney +11
The endowment of AI with reasoning capabilities and some degree of agency is widely viewed as a path toward more capable and generalizable systems. Our position is that the current…
Protecting Users From Themselves: Safeguarding Contextual Privacy in Interactions with Conversational Agents
Ivoline Ngong, Swanand Kadhe, Hao Wang +4
Conversational agents are increasingly woven into individuals' personal lives, yet users often underestimate the privacy risks associated with them. The moment users share informat…
Final-Model-Only Data Attribution with a Unifying View of Gradient-Based Methods
Dennis Wei, Inkit Padhi, Soumya Ghosh +3
Training data attribution (TDA) is concerned with understanding model behavior in terms of the training data. This paper draws attention to the common setting where one has access…
Identifying Sub-networks in Neural Networks via Functionally Similar Representations
Tian Gao, Amit Dhurandhar, Karthikeyan Natesan Ramamurthy +1
Providing human-understandable insights into the inner workings of neural networks is an important step toward achieving more explainable and trustworthy AI. Existing approaches to…
Programming Refusal with Conditional Activation Steering
Bruce W. Lee, Inkit Padhi, Karthikeyan Natesan Ramamurthy +4
LLMs have shown remarkable capabilities, but precisely controlling their response behavior remains challenging. Existing activation steering methods alter LLM behavior indiscrimina…
CELL your Model: Contrastive Explanations for Large Language Models
Ronny Luss, Erik Miehling, Amit Dhurandhar
The advent of black-box deep neural network classification models has sparked the need to explain their decisions. However, in the case of generative AI, such as large language mod…
Large Language Model Confidence Estimation via Black-Box Access
Tejaswini Pedapati, Amit Dhurandhar, Soumya Ghosh +2
Estimating uncertainty or confidence in the responses of a model can be significant in evaluating trust not only in the responses, but also in the model as a whole. In this paper,…
NeuroPrune: A Neuro-inspired Topological Sparse Training Algorithm for Large Language Models
Amit Dhurandhar, Tejaswini Pedapati, Ronny Luss +4
Transformer-based Language Models have become ubiquitous in Natural Language Processing (NLP) due to their impressive performance on various tasks. However, expensive training as w…
Multi-Level Explanations for Generative Language Models
Lucas Monteiro Paes, Dennis Wei, Hyo Jin Do +8
Despite the increasing use of large language models (LLMs) for context-grounded tasks like summarization and question-answering, understanding what makes an LLM produce a certain r…
Ranking Large Language Models without Ground Truth
Amit Dhurandhar, Rahul Nair, Moninder Singh +2
Evaluation and ranking of large language models (LLMs) has become an important problem with the proliferation of these models and their impact. Evaluation methods either require hu…
Trust Regions for Explanations via Black-Box Probabilistic Certification
Amit Dhurandhar, Swagatam Haldar, Dennis Wei +1
Given the black box nature of machine learning models, a plethora of explainability methods have been developed to decipher the factors behind individual decisions. In this paper,…