From 72b4076d39e357348c205b51cba2f20faf91be8f Mon Sep 17 00:00:00 2001 From: xiaoyang Date: Sat, 25 Jul 2026 19:08:29 +0800 Subject: [PATCH 1/3] docs: update search-agent promotion copy --- README.md | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/README.md b/README.md index 57864c01a..81b1679ca 100644 --- a/README.md +++ b/README.md @@ -31,6 +31,7 @@ # 📰 News | 🗞️ News | 📝 Description | | -- | ------ | +| Search Agent Post-Training | We applied the R&D-Agent method to the search-agent setting, covering the full pipeline end to end. The open-sourced building blocks are [axrl](https://github.com/XYZ-AI-Lab/axrl) for infra and [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness. | | ICML 2026 Acceptance | We are thrilled to announce that our paper [FT-Dojo: Towards Autonomous LLM Fine-Tuning with Language Agents](https://arxiv.org/abs/2603.01712) has been accepted to ICML 2026. The FT-Agent implementation is available in the [LLM fine-tuning guide](rdagent/app/finetune/llm/README.md). | | ACL 2026 Findings Acceptance | We are thrilled to announce that our paper [Reasoning as Gradient](https://arxiv.org/abs/2603.01692) has been accepted to ACL 2026 Findings. Execution traces are available at [Gome GPT-5 Traces](https://huggingface.co/datasets/amstrongzyf/Gome-GPT5-Traces) | | Web UI Release | We release a new frontend that can be built and served by `rdagent server_ui` for real-time interaction and trace viewing, currently excluding the `data_science` scenario. | @@ -71,6 +72,16 @@ You can inspect the detailed runs of the above results online. For running R&D-Agent on MLE-bench, refer to **[MLE-bench Guide: Running ML Engineering via MLE-bench](https://rdagent.readthedocs.io/en/latest/scens/data_science.html)** +# 🔎 Search Agent Post-Training + +We applied the R&D-Agent method to the search-agent setting and extended it to cover the full pipeline end to end. +The two open-sourced building blocks are: + +- [axrl](https://github.com/XYZ-AI-Lab/axrl) for the infra layer +- [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness layer + +Together, they carry RD-Agent into the post-training stage of search agents. + # 🥇 The First Data-Centric Quant Multi-Agent Framework! R&D-Agent for Quantitative Finance, in short **RD-Agent(Q)**, is the first data-centric, multi-agent framework designed to automate the full-stack research and development of quantitative strategies via coordinated factor-model co-optimization. From 3e3b6ad92020f3c65d00e872abb25adc644ae314 Mon Sep 17 00:00:00 2001 From: xiaoyang Date: Sat, 25 Jul 2026 20:12:48 +0800 Subject: [PATCH 2/3] docs: highlight XYZ AI Lab search-agent stack --- README.md | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index 81b1679ca..590e12144 100644 --- a/README.md +++ b/README.md @@ -31,7 +31,7 @@ # 📰 News | 🗞️ News | 📝 Description | | -- | ------ | -| Search Agent Post-Training | We applied the R&D-Agent method to the search-agent setting, covering the full pipeline end to end. The open-sourced building blocks are [axrl](https://github.com/XYZ-AI-Lab/axrl) for infra and [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness. | +| Search Agent Post-Training | Together with [XYZ AI Lab](https://xyz-lab.ai/), we applied the R&D-Agent method to the search-agent setting, covering the full pipeline end to end. The open-sourced building blocks are [axrl](https://github.com/XYZ-AI-Lab/axrl) for infra and [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness. | | ICML 2026 Acceptance | We are thrilled to announce that our paper [FT-Dojo: Towards Autonomous LLM Fine-Tuning with Language Agents](https://arxiv.org/abs/2603.01712) has been accepted to ICML 2026. The FT-Agent implementation is available in the [LLM fine-tuning guide](rdagent/app/finetune/llm/README.md). | | ACL 2026 Findings Acceptance | We are thrilled to announce that our paper [Reasoning as Gradient](https://arxiv.org/abs/2603.01692) has been accepted to ACL 2026 Findings. Execution traces are available at [Gome GPT-5 Traces](https://huggingface.co/datasets/amstrongzyf/Gome-GPT5-Traces) | | Web UI Release | We release a new frontend that can be built and served by `rdagent server_ui` for real-time interaction and trace viewing, currently excluding the `data_science` scenario. | @@ -74,13 +74,19 @@ For running R&D-Agent on MLE-bench, refer to **[MLE-bench Guide: Running ML Engi # 🔎 Search Agent Post-Training -We applied the R&D-Agent method to the search-agent setting and extended it to cover the full pipeline end to end. -The two open-sourced building blocks are: +

+ + XYZ AI Lab + +

+ +Together with [XYZ AI Lab](https://xyz-lab.ai/), we applied the R&D-Agent method to the search-agent setting and extended it to cover the full post-training pipeline end to end. +The stack is now open through two complementary projects: - [axrl](https://github.com/XYZ-AI-Lab/axrl) for the infra layer - [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness layer -Together, they carry RD-Agent into the post-training stage of search agents. +Together, they bring RD-Agent into search-agent post-training with a more complete path from experimentation to evaluation. # 🥇 The First Data-Centric Quant Multi-Agent Framework! From 06a502b435ea7f127bf034735870fbd2df7c96ab Mon Sep 17 00:00:00 2001 From: xiaoyang Date: Sat, 25 Jul 2026 20:41:17 +0800 Subject: [PATCH 3/3] docs: add XYZ AI Lab branding to promo copy --- README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index 590e12144..2ac9374e0 100644 --- a/README.md +++ b/README.md @@ -31,7 +31,7 @@ # 📰 News | 🗞️ News | 📝 Description | | -- | ------ | -| Search Agent Post-Training | Together with [XYZ AI Lab](https://xyz-lab.ai/), we applied the R&D-Agent method to the search-agent setting, covering the full pipeline end to end. The open-sourced building blocks are [axrl](https://github.com/XYZ-AI-Lab/axrl) for infra and [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness. | +| XYZ AI Lab Search Agent Post-Training | Together with [XYZ AI Lab](https://xyz-lab.ai/), we applied the R&D-Agent method to the search-agent setting, covering the full pipeline end to end. The open-sourced building blocks are :octocat: [axrl](https://github.com/XYZ-AI-Lab/axrl) for infra and :octocat: [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness. | | ICML 2026 Acceptance | We are thrilled to announce that our paper [FT-Dojo: Towards Autonomous LLM Fine-Tuning with Language Agents](https://arxiv.org/abs/2603.01712) has been accepted to ICML 2026. The FT-Agent implementation is available in the [LLM fine-tuning guide](rdagent/app/finetune/llm/README.md). | | ACL 2026 Findings Acceptance | We are thrilled to announce that our paper [Reasoning as Gradient](https://arxiv.org/abs/2603.01692) has been accepted to ACL 2026 Findings. Execution traces are available at [Gome GPT-5 Traces](https://huggingface.co/datasets/amstrongzyf/Gome-GPT5-Traces) | | Web UI Release | We release a new frontend that can be built and served by `rdagent server_ui` for real-time interaction and trace viewing, currently excluding the `data_science` scenario. | @@ -83,8 +83,8 @@ For running R&D-Agent on MLE-bench, refer to **[MLE-bench Guide: Running ML Engi Together with [XYZ AI Lab](https://xyz-lab.ai/), we applied the R&D-Agent method to the search-agent setting and extended it to cover the full post-training pipeline end to end. The stack is now open through two complementary projects: -- [axrl](https://github.com/XYZ-AI-Lab/axrl) for the infra layer -- [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness layer +- :octocat: [axrl](https://github.com/XYZ-AI-Lab/axrl) for the infra layer +- :octocat: [AxisAgentic](https://github.com/XYZ-AI-Lab/AxisAgentic) for the harness layer Together, they bring RD-Agent into search-agent post-training with a more complete path from experimentation to evaluation.