6 citations · 10 across the 17 of their papers we have counts for
17 papers
When Policies Change Probabilities: Modular Decision-Making for LLM Code Review
Rasvik Kudum, Max Corbett, Hitansh Paliwal +3
LLM code reviewers often estimate patch risk and make approval decisions in one prompt. A probability should depend on evidence; costs should determine the action taken from it. We…
CIAware-Bench: Benchmarking Control Intervention Awareness Across Frontier LLMs
Joachim Schaeffer, Thomas Jiralerspong, Alexander Panfilov +4
AI control protocols oversee untrusted models by monitoring their actions and modifying potentially unsafe steps, often using a trusted model. This partially tampers with the untru…
Noticing the Watcher: LLM Agents Can Infer CoT Monitoring from Blocking Feedback
Thomas Jiralerspong, Flemming Kondrup, Yoshua Bengio
Chain-of-thought (CoT) monitoring is proposed as a method for overseeing the internal reasoning of language-model agents. Prior work has shown that when models are explicitly infor…
Delta-Crosscoder: Robust Crosscoder Model Diffing in Narrow Fine-Tuning Regimes
Aly Kassem, Thomas Jiralerspong, Negar Rostamzadeh +1
Model diffing methods aim to identify how fine-tuning changes a model's internal representations. Crosscoders approach this by learning shared dictionaries of interpretable latent…
Cross-Architecture Model Diffing with Crosscoders: Unsupervised Discovery of Differences Between LLMs
Thomas Jiralerspong, Trenton Bricken
Model diffing, the process of comparing models' internal representations to identify their differences, is a promising approach for uncovering safety-critical behaviors in new mode…
Learning What Matters: Steering Diffusion via Spectrally Anisotropic Forward Noise
Luca Scimeca, Thomas Jiralerspong, Berton Earnshaw +2
Diffusion Probabilistic Models (DPMs) have achieved strong generative performance, yet their inductive biases remain largely implicit. In this work, we aim to build inductive biase…
Shaping Inductive Bias in Diffusion Models through Frequency-Based Noise Control
Thomas Jiralerspong, Berton Earnshaw, Jason Hartford +2
Diffusion Probabilistic Models (DPMs) are powerful generative models that have achieved unparalleled success in a number of generative tasks. In this work, we aim to build inductiv…
General Causal Imputation via Synthetic Interventions
Marco Jiralerspong, Thomas Jiralerspong, Vedant Shah +2
Given two sets of elements (such as cell types and drug compounds), researchers typically only have access to a limited subset of their interactions. The task of causal imputation…
A Complexity-Based Theory of Compositionality
Eric Elmoznino, Thomas Jiralerspong, Yoshua Bengio +1
Compositionality is believed to be fundamental to intelligence. In humans, it underlies the structure of thought, language, and higher-level reasoning. In AI, compositional represe…
Geometric Signatures of Compositionality Across a Language Model's Lifetime
Jin Hwa Lee, Thomas Jiralerspong, Lei Yu +2
By virtue of linguistic compositionality, few syntactic rules and a finite lexicon can generate an unbounded number of sentences. That is, language, though seemingly high-dimension…
Expressivity of Neural Networks with Random Weights and Learned Biases
Ezekiel Williams, Alexandre Payeur, Avery Hee-Woon Ryoo +4
Landmark universal function approximation results for neural networks with trained weights and biases provided the impetus for the ubiquitous use of neural networks as learning mod…
Efficient Causal Graph Discovery Using Large Language Models
Thomas Jiralerspong, Xiaoyin Chen, Yash More +2
We propose a novel framework that leverages LLMs for full causal graph discovery. While previous LLM-based methods have used a pairwise query approach, this requires a quadratic nu…
Network Analysis of the iNaturalist Citizen Science Community
Yu Lu Liu, Thomas Jiralerspong
In recent years, citizen science has become a larger and larger part of the scientific community. Its ability to crowd source data and expertise from thousands of citizen scientist…
Forecaster: Towards Temporally Abstract Tree-Search Planning from Pixels
Thomas Jiralerspong, Flemming Kondrup, Doina Precup +1
The ability to plan at many different levels of abstraction enables agents to envision the long-term repercussions of their decisions and thus enables sample-efficient learning. Th…
A Comparison of Classical and Deep Reinforcement Learning Methods for HVAC Control
Marshall Wang, John Willes, Thomas Jiralerspong +1
Reinforcement learning (RL) is a promising approach for optimizing HVAC control. RL offers a framework for improving system performance, reducing energy consumption, and enhancing…
Towards Safe Mechanical Ventilation Treatment Using Deep Offline Reinforcement Learning
Flemming Kondrup, Thomas Jiralerspong, Elaine Lau +5
Mechanical ventilation is a key form of life support for patients with pulmonary impairment. Healthcare workers are required to continuously adjust ventilator settings for each pat…
Contrastive Retrospection: honing in on critical steps for rapid learning and generalization in RL
Chen Sun, Wannan Yang, Thomas Jiralerspong +4
In real life, success is often contingent upon multiple critical steps that are distant in time from each other and from the final reward. These critical steps are challenging to i…