ZARZA We are Zarza, the prestigious firm behind major projects in information technology.

Optimizing Test-Time Compute via Meta Reinforcement Fine-Tuning

Best AI papers explained

14/03/2025 11 min

Listen "Optimizing Test-Time Compute via Meta Reinforcement Fine-Tuning"

Descargar episodio Ver en sitio original

Episode Synopsis

Longer version

More episodes of the podcast Best AI papers explained

PRL: Process Reward Learning Improves LLMs’ Reasoning Ability and Broadens the Reasoning Boundary 18/01/2026

Coverage Improvement and Fast Convergence of On-policy Preference Learning 17/01/2026

Stagewise Reinforcement Learning and the Geometry of the Regret Landscape 16/01/2026

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models 16/01/2026

Learning Latent Action World Models In The Wild 16/01/2026

From Unstructured Data to Demand Counterfactuals: Theory and Practice 14/01/2026

In-context reinforcement learning through bayesian fusion of context and value prior 14/01/2026

Digital RedQueen: Adversarial Program Evolution in Core War with LLMs 14/01/2026

Extending the Context of Pretrained LLMs by Dropping Their Positional Embeddings 13/01/2026

Representation-Based Exploration for Language Models: from test-time to post-training 12/01/2026

Ver todos los episodios