From c6d33c2a0dd2989eb4886f828a556c1ec6f5d17a Mon Sep 17 00:00:00 2001
From: genquan9 <49327371+genquan9@users.noreply.github.com>
Date: Wed, 5 Aug 2026 16:18:11 -0700
Subject: [PATCH] Add tau2-bench RL blog post to docs README (#4561)

---
 docs/README.md | 1 +
 1 file changed, 1 insertion(+)

diff --git a/docs/README.md b/docs/README.md
index 59b59fedf..684ebf4ac 100644
--- a/docs/README.md
+++ b/docs/README.md
@@ -8,3 +8,4 @@ This directory contains deep-dive documentation, extended guides, and architectu
 ## Blog Posts
 - [Model Distillation Best Practices](https://googlecloudplatform.github.io/vertex-ai-samples/vertex-training-cluster/model_distillation_best_practices): Explores off-policy model distillation, dataset curation, and hyperparameter scaling laws for training student models on Vertex AI.
 - [Forgetting Mitigation via Data Mixing](https://googlecloudplatform.github.io/vertex-ai-samples/vertex-training-cluster/forgetting_mitigation_data_mixing): Discusses catastrophic forgetting in model fine-tuning and how to mitigate it using multi-domain data mixing on Vertex AI.
+- [Multi-Turn Reinforcement Learning for τ²-bench](https://googlecloudplatform.github.io/vertex-ai-samples/vertex-training-cluster/multi_turn_reinforcement_learning_for_tau2_bench): Explores multi-turn RL training for tool-calling agents using GRPO on the τ²-bench customer service benchmark with NeMo RL.