10 papers
Self-Play Pretraining with Zero Data
Aditya Cowsik, Kfir Dolev, Michael Y. Li +4
Advances in language modeling have been driven by scaling pretraining on ever more data. Yet, the training data is still largely curated on the model's behalf. A more general appro…
QuasiMoTTo: Quasi-Monte Carlo Test-Time Scaling
Michael Y. Li, Anthony Zhan, Kanishk Gandhi +2
Scaling inference compute, by generating many parallel attempts per problem, is a costly but reliable lever for improving language model capabilities. By default these attempts are…
auto-psych: Automating the science of mind using agent-driven theory discovery and experimentation
Ben Prystawski, Kushin Mukherjee, Daniel Wurgaft +4
AI-based scientific automation is increasingly possible by using agents to generate hypotheses, design experiments, and analyze data. Data collection is a major bottleneck in this…
SPIRAL: Learning to Search and Aggregate
Jubayer Ibn Hamid, Ifdita Hasan Orney, Michael Y. Li +5
Language model reasoning can be substantially improved at test time via scaffolds that scale inference compute across different primitives -- sequential reasoning within a trace, i…
CORE: Contrastive Reflection Enables Rapid Improvements in Reasoning
Linas Nasvytis, Simon Jerome Han, Ben Prystawski +3
Language models can use verifiable rewards to improve at a wide variety of reasoning tasks. However, both parametric (e.g. RLVR) and non-parametric (e.g. prompt optimization) appro…
Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior
Daniel Wurgaft, Can Rager, Matthew Kowal +13
Neural representations carry rich geometric structure; but does that structure causally shape behavior? To address this question, we intervene along paths through activation space…
Poly-EPO: Training Exploratory Reasoning Models
Ifdita Hasan Orney, Jubayer Ibn Hamid, Shreya S Ramanujam +5
Exploration is a cornerstone of learning from experience: it enables agents to find solutions to complex problems, generalize to novel ones, and scale performance with test-time co…
Neural Garbage Collection: Learning to Forget while Learning to Reason
Michael Y. Li, Jubayer Ibn Hamid, Emily B. Fox +1
Chain-of-thought reasoning has driven striking advances in language model capability, yet every reasoning step grows the KV cache, creating a bottleneck to scaling this paradigm fu…
Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering
Eric Bigelow, Daniel Wurgaft, YingQiao Wang +4
Large language models (LLMs) can be controlled at inference time through prompts (in-context learning) and internal activations (activation steering). Different accounts have been…
In-Context Learning Strategies Emerge Rationally
Daniel Wurgaft, Ekdeep Singh Lubana, Core Francisco Park +3
Recent work analyzing in-context learning (ICL) has identified a broad set of strategies that describe model behavior in different experimental conditions. We aim to unify these fi…