5 citations · 15 across the 18 of their papers we have counts for
20 papers
Effectiveness of Hidden-State Probes for Localizing Reasoning Failures
Shubham Subhnil, Yannis Belkhiter, Erik Miehling +2
Verifying the final answer of a long reasoning trace may show that the trace is wrong but it does not distinguish three cases. The problem may lie beyond the model's capability, an…
On the Steering Dimensionality of Refusal in Language Models
Han Wang, Erik Miehling, Dennis Wei +2
Existing activation steering methods often assume that a high-level concept can be mediated by a single steering direction. To support this, two complementary interventions should…
Local Causal Attribution of Chain-of-Thought Reasoning
Dennis Wei, Yannis Belkhiter, Erik Miehling +1
Understanding the causal structure of a language model's thought process is a problem of significant importance for both transparency and safety. In this work, we take a local appr…
AI Steerability 360: A Toolkit for Steering Large Language Models
Erik Miehling, Karthikeyan Natesan Ramamurthy, Praveen Venkateswaran +10
The AI Steerability 360 toolkit is an extensible, open-source Python library for steering LLMs. Steering abstractions are designed around four model control surfaces: input (modifi…
The Effectiveness of Approximate Regularized Replay for Efficient Supervised Fine-Tuning of Large Language Models
Matthew Riemer, Erik Miehling, Miao Liu +2
Although parameter-efficient fine-tuning methods, such as LoRA, only modify a small subset of parameters, they can have a significant impact on the model. Our instruction-tuning ex…
ICX360: In-Context eXplainability 360 Toolkit
Dennis Wei, Ronny Luss, Xiaomeng Hu +6
Large Language Models (LLMs) have become ubiquitous in everyday life and are entering higher-stakes applications ranging from summarizing meeting transcripts to answering doctors'…
Generate, Evaluate, Iterate: Synthetic Data for Human-in-the-Loop Refinement of LLM Judges
Hyo Jin Do, Zahra Ashktorab, Jasmina Gajcin +5
The LLM-as-a-judge paradigm enables flexible, user-defined evaluation, but its effectiveness is often limited by the scarcity of diverse, representative data for refining criteria.…
Who Sees the Risk? Stakeholder Conflicts and Explanatory Policies in LLM-based Risk Assessment
Srishti Yadav, Jasmina Gajcin, Erik Miehling +1
Understanding how different stakeholders perceive risks in AI systems is essential for their responsible deployment. This paper presents a framework for stakeholder-grounded risk a…
Interpreting LLM-as-a-Judge Policies via Verifiable Global Explanations
Jasmina Gajcin, Erik Miehling, Rahul Nair +3
Using LLMs to evaluate text, that is, LLM-as-a-judge, is increasingly being used at scale to augment or even replace human annotations. As such, it is imperative that we understand…
Language Models Coupled with Metacognition Can Outperform Reasoning Models
Vedant Khandelwal, Francesca Rossi, Keerthiram Murugesan +4
Large language models (LLMs) excel in speed and adaptability across various reasoning tasks, but they often struggle when strict logic or constraint enforcement is required. In con…
EvalAssist: A Human-Centered Tool for LLM-as-a-Judge
Zahra Ashktorab, Werner Geyer, Michael Desmond +6
With the broad availability of large language models and their ability to generate vast outputs using varied prompts and configurations, determining the best output for a given tas…
Localizing Persona Representations in LLMs
Celia Cintas, Miriam Rateike, Erik Miehling +2
We present a study on how and where personas -- defined by distinct sets of human characteristics, values, and beliefs -- are encoded in the representation space of large language…
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…
Granite Guardian
Inkit Padhi, Manish Nagireddy, Giandomenico Cornacchia +20
We introduce the Granite Guardian models, a suite of safeguards designed to provide risk detection for prompts and responses, enabling safe and responsible use in combination with…
Evaluating the Prompt Steerability of Large Language Models
Erik Miehling, Michael Desmond, Karthikeyan Natesan Ramamurthy +5
Building pluralistic AI requires designing models that are able to be shaped to represent a wide range of value systems and cultures. Achieving this requires first being able to ev…
Attack Atlas: A Practitioner's Perspective on Challenges and Pitfalls in Red Teaming GenAI
Ambrish Rawat, Stefan Schoepf, Giulio Zizzo +10
As generative AI, particularly large language models (LLMs), become increasingly integrated into production applications, new attack surfaces and vulnerabilities emerge and put a f…
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…
Language Models in Dialogue: Conversational Maxims for Human-AI Interactions
Erik Miehling, Manish Nagireddy, Prasanna Sattigeri +3
Modern language models, while sophisticated, exhibit some inherent shortcomings, particularly in conversational settings. We claim that many of the observed shortcomings can be att…
Detectors for Safe and Reliable LLMs: Implementations, Uses, and Limitations
Swapnaja Achintalwar, Adriana Alvarado Garcia, Ateret Anaby-Tavor +35
Large language models (LLMs) are susceptible to a variety of risks, from non-faithful output to biased and toxic generations. Due to several limiting factors surrounding LLMs (trai…