AuraTracer智迹闻
中文

EVENT DOSSIER

Iris: Climbing to the Search Frontier

The Iris team released the Iris-mini and Iris-pro models, as well as the SFT-RL c… on September 7, 2026.

2026-09-07 12:00 Models 🔥 53.2 heat score hf-papers #19
2sources
1days unfolding
53.2heat score
3mentions
SummaryAI generated

On September 7, 2026, the Iris team released two search proxy models: Iris-mini and Iris-pro. Iris-mini was trained on a dataset of 35B-A3B size, while Iris-pro was trained on a dataset of 397B-A17B size. The study proposed a training process called “SFT-RL climbing”: first, tasks were constructed back from the web hyperlink structure and data was filtered; after supervised fine-tuning (SFT), reinforcement learning (RL) optimization was performed using a reward discriminator and an observation summarizer within the training cluster. Ultra-long inferences were interrupted at the request layer to retain promised prefixes. Evaluation included benchmarks such as BrowseComp, BrowseComp-ZH, DeepSearchQA, and HLE. All results came from a single ReAct proxy, and context management functions were enabled when inference was performed. With this configuration, Iris-pro achieved scores of 88.6/85.1/92.9/56.4 on the four benchmarks…

Related eventsRELATED EVENTS
Quick factsQUICK FACTS
35B-A3BModel size
397B-A17BModel size
Key entitiesKEY ENTITIES
Hugging FaceIris-miniIris-pro

Event frameEVENT FRAME

Launch

政府 · 科研机构across 1 days

Status

The Iris team released the Iris-mini and Iris-pro models, as well as the SFT-RL c… on September 7, 2026.

Coverage · reports per dayLANGUAGE SPLIT

Entity relations
Iris-mini × Iris-pro2Hugging Face × Iris-mini1Hugging Face × Iris-pro1

Integrated timelineUNIFIED TIMELINE

  1. 2026-09-07

    The Iris team released the search proxy model.

    The Iris team released Iris-mini and Iris-pro, trained on models with sizes of 35B-A3B and 397B-A17B respectively, and proposed the SFT-RL climbing process.

    2 reports

SignalsSIGNALS

Keyword heat
  • Iris-mini2
  • Iris-pro2
  • Hugging Face1

All reports (2)SOURCES

H Hugging Face Papers en 2026-09-07 08:00

Paper page - Iris: Climbing to the Search Frontier

Hugging Face has released two search proxy models, Iris-mini and Iris-pro, trained respectively on 35B-A3B and 397B-A17B datasets. This study proposes the “SFT-RL climbing” process: tasks are constructed reversely from web hyperlink structures, problems are filtered and optimized through reinforcement learning, and the most difficult and efficient trajectories are returned to the supervised fine-tuning phase. With context management enabled during inference, Iris-pro achieved scores of 88.6/85.1/92.9/56.4 on the BrowseComp, BrowseComp-ZH, DeepSearchQA, and HLE benchmarks, making it the strongest open-source search proxy in its class in terms of parameter size. The authors plan to release the complete data construction, training, and evaluation recipes along with the model weights.

A arXiv cs.AI en 2026-09-07 12:00

Iris: Climbing to the Search Frontier

The Iris team has released two search agents, Iris-mini and Iris-pro, trained on datasets of 35B-A3B and 397B-A17B respectively. The study proposes the SFT-RL climbing process: tasks are constructed backward from the web hyperlink structure, data is filtered at the trajectory and round levels through supervised fine-tuning (SFT), followed by reinforcement learning (RL) optimization using a reward classifier and an observation summarizer within the training cluster, and long-range reasoning is interrupted at the request layer to retain promised prefixes. Evaluations included BrowseComp, BrowseComp-ZH, DeepSearchQA, and HLE benchmarks, with all results coming from a single ReAct agent and context management enabled when reasoning was active. With this feature enabled, Iris-pro achieved scores of 88.6/85.1/92.9/56.4 on the four benchmarks, making it the most performant open-source search agent in terms of respective parameter sizes. The team plans to release model weights and the complete…