SIREN, a selective entropy regularization method, boosted the Qwen2.5‑Math‑7B model by 6.6 points on the majority‑at‑k metric for AIME24/25 benchmarks. Read more: https://getnews.me/selective-entropy-regularization-boosts-large-reasoning-models/ #siren #largereasoningmodels #entropyregularization

The paper adds an L₂‑entropy term to optimal transport, creating a finite‑dimensional particle control method that yields uniformly sampled reachable sets. https://getnews.me/entropy-regularized-method-for-uniform-sampling-of-reachable-sets/ #entropyregularization #optimaltransport #reachablesets