PulseAugur
EN
LIVE 05:41:57

Iris search agents achieve new frontier results on web benchmarks

By PulseAugur Editorial · [2 sources] ·

Researchers have introduced Iris-mini and Iris-pro, two search agents trained at 35B-A3B and 397B-A17B scales, respectively. These agents are developed using a novel data pipeline that constructs multi-hop chains over web corpus entity graphs and generates questions that are challenging for reference models. The training process involves alternating supervised fine-tuning (SFT) with reinforcement learning (RL) against live search, a method termed SFT-RL climbing. When evaluated with context management, Iris-pro achieved strong results on benchmarks like BrowseComp and DeepSearchQA, outperforming other open-source search agents in its parameter range. AI

IMPACT These models advance the capabilities of open-source search agents, potentially improving web navigation and information retrieval.

RANK_REASON The item is a research paper detailing new AI models and training methodologies. [lever_c_demoted from research: ic=1 ai=1.0]

Read on Hugging Face Daily Papers →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

Iris search agents achieve new frontier results on web benchmarks

How we ranked this

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The item is a research paper detailing new AI models and training methodologies. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
2 independent sources
Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics
model release, paper, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
4 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.
Coverage growth since scoring
+1 source(s) since last score
New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

Full methodology in our editorial standards.

COVERAGE [2]

  1. arXiv cs.AI TIER_1 English(EN) · Ziyuan Liu, Hengqi Liu, Zichuan Wang, Yang Qin, Jiachen Liang, Xu Chu, Shaowei Chen, Yuantao Gu, Mu Chuan ·

    Iris: Climbing to the Search Frontier

    arXiv:2609.04304v1 Announce Type: new Abstract: We present Iris-mini and Iris-pro, two search agents trained at the 35B-A3B and 397B-A17B scales, together with the data pipeline and training recipe behind them. Tasks are reverse-constructed from the hyperlink structure of a web c…

  2. Hugging Face Daily Papers TIER_1 English(EN) ·

    Iris: Climbing to the Search Frontier

    Two large-scale search agents are trained via a multi-stage pipeline combining supervised fine-tuning and reinforcement learning against live search, achieving state-of-the-art open-source results on complex web benchmarks through rigorous trajectory filtering and inference-time …

RELATED ENTITIES

RELATED TOPICS