T-Router: Learning Thalamic Routing for Reasoning with Parameter-Efficient Reinforcement Learning

概要

Parameter-efficient reinforcement learning aims to improve reasoning with a compact trainable interface to a pretrained model. We introduce the Thalamic Router (T-Router), which concentrates adaptation on the reuse of completed computations. A compressed, addressable bank preserves block changes; a depth-recurrent controller conditions their selection and relative-scale writeback. This coupling gives thalamic context-dependent routing a concrete computational form: learn which earlier contributions a receiving layer uses, and with what influence. Correctness rewards train the interface while preserving backbone parameters and layer order. On an 8.95B-parameter backbone, T-Router allocates 41.73M parameters (0.466% of the backbone) and achieves 83.64 +/- 1.16 MathAvg after GSM8K RL, compared with 73.79 +/- 1.83 for full-parameter GRPO across three evaluation rounds. At a comparable parameter budget and with matched retries, it exceeds LoRA's 77.28 +/- 1.95 MathAvg, improving all three task families and raising mean AIME accuracy from 48.33 to 60.56. Capacity-controlled comparisons favor addressable block changes and recurrent context; separate search training extends the interface to tool-mediated reasoning. These results establish controlled computation reuse as an effective route to parameter-efficient reasoning reinforcement learning.

参考リンク

• Paper ↗

リンク先のライセンス・再現性・データ条件をご確認ください。自動分類や実装の目安は保証ではありません。

実装を始めるプロンプト

参考リンクを確認したうえで、次の内容をコピーして利用できます。

# 論文・記事
T-Router: Learning Thalamic Routing for Reasoning with Parameter-Efficient Reinforcement Learning

# 概要
Parameter-efficient reinforcement learning aims to improve reasoning with a compact trainable interface to a pretrained model. We introduce the Thalamic Router (T-Router), which concentrates…

# 原典
https://arxiv.org/abs/2609.39109

上記の手法を実装してください。実行前に依存関係とライセンスを確認してください。