collaborators

6 papers

cs.CL2026

Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment

Cameron Tice, Puria Radmard, Samuel Ratnam +3

Pretraining corpora contain extensive discourse about AI systems, yet the causal influence of this discourse on downstream alignment remains poorly understood. If prevailing descri…

cs.LG2026

Deep Ignorance: Filtering Pretraining Data Builds Tamper-Resistant Safeguards into Open-Weight LLMs

Kyle O'Brien, Stephen Casper, Quentin Anthony +7

Open-weight AI systems offer unique benefits, including enhanced transparency, open research, and decentralized access. However, they are vulnerable to tampering attacks which can…

cs.CY2026

Expanding External Access To Frontier AI Models For Dangerous Capability Evaluations

Jacob Charnock, Alejandro Tlaie, Kyle O'Brien +2

Frontier AI companies increasingly rely on external evaluations to assess risks from dangerous capabilities before deployment. However, external evaluators often receive limited mo…

cs.LG2025

Steering Language Model Refusal with Sparse Autoencoders

Kyle O'Brien, David Majercak, Xavier Fernandes +7

Responsible deployment of language models requires mechanisms for refusing unsafe prompts while preserving model performance. While most approaches modify model weights through add…

cs.CL2025

Recite, Reconstruct, Recollect: Memorization in LMs as a Multifaceted Phenomenon

USVSN Sai Prashanth, Alvin Deng, Kyle O'Brien +9

Memorization in language models is typically treated as a homogenous phenomenon, neglecting the specifics of the memorized data. We instead model memorization as the effect of a se…

cs.LG2025

Composable Interventions for Language Models

Arinbjorn Kolbeinsson, Kyle O'Brien, Tianjin Huang +8

Test-time interventions for language models can enhance factual accuracy, mitigate harmful outputs, and improve model efficiency without costly retraining. But despite a flood of n…