Ploba logo

Find the right AI tool for the job: apps, agents, APIs, MCP servers and datasets, explained plainly.

Get the AI Digest

Discover

  • Explore
  • Top Charts
  • Categories
  • Agents
  • MCP servers
  • Developer tools

Learn & community

  • Learn AI
  • AI Glossary
  • Blog
  • Hackathons
  • People
  • AI Digest

For makers

  • List your tool
  • Docs
  • Claim a listing
  • Advertise
  • Write for us

Company

  • About
  • Careers
  • Changelog
  • Trust
  • Security
  • Contact

© 2026 Ploba. All rights reserved.

  • Privacy
  • Terms
  • Cookies
  • Advertising
  • All policies
llms.txt
Ploba logo
DiscoverHackathonsPeopleLearnAI Digest
Sign inGet started
HomeTop ChartsExploreCategories
TypesWeb apps1,894Agents585Plugins416APIs376Mobile apps152Datasets85
AdvertisingYour product hereReach people browsing AI tools on Ploba.Advertise hereAdvertisingYour product hereReach people browsing AI tools on Ploba.Advertise here
  1. Explore
  2. /
  3. AI
  4. /
  5. UltraData RL 2609
AI
UR

UltraData RL 2609

by OpenBMB

UltraData-RL-2609 is OpenBMB's reinforcement learning dataset of about 86,000 verifiable tasks across math, science, long-context QA and code.

Dataset·Free

View on Hugging Face(opens in a new tab)

About UltraData RL 2609

Key features of UltraData RL 2609

  • 85,995 verifiable RL training samples
  • Four domains: math, code, long-context and knowledge
  • Answer matching for math, knowledge and long-context tasks
  • Code checked by running submissions against test cases
  • Label checks using an LLM judge and multi-model consensus
  • Difficulty calibrated to the RL starting model
  • JSONL with five fields per sample

Who UltraData RL 2609 is for

  • Researchers doing RL on reasoning models
  • Teams post-training small language models
  • Developers building verifiable-reward training pipelines

How to use UltraData RL 2609

  1. 1

    Step 1: Install the datasets library with pip install datasets.

  2. 2

    Step 2: Load the data with load_dataset("openbmb/UltraData-RL-2609").

  3. 3

    Step 3: Filter by the domain field to pick Math, Knowledge, Long_Context or Code.

  4. 4

    Step 4: Use query as the prompt and ground_truth as the reward target.

  5. 5

    Step 5: For code samples, pipe each test input to the candidate program and compare stdout with the expected output.

UltraData RL 2609 FAQ

What is in UltraData-RL-2609?

85,995 verifiable tasks: math, code, long-context QA and knowledge (STEM) questions, each with a checkable reference answer or test cases.

How are code answers checked?

By execution. A submission is correct only if its stdout matches the expected output for every test case.

What fields does each sample have?

uuid, query, ground_truth, source and domain.

Which tasks were removed?

Multiple-choice, true/false, proof, multi-part and image-dependent items were removed during construction.

Details

Pricing
Free
Type
Dataset
Categories
AIDataResearch
Added
Sep 17, 2026
OpPublisherOpenBMB

Built from public sources

We put this page together from UltraData RL 2609’s own website and docs so you can compare it with other tools. Ploba isn’t affiliated with UltraData RL 2609.

Work at UltraData RL 2609? Claim the listing to manage it yourself.

Something wrong here?

More from OpenBMB

View publisher
UlUltra-FineWebOpenBMB · FreeUltra-FineWeb is OpenBMB's filtered web pre-training dataset, with about 1T English tokens and 120B Chinese tokens. It is built by applying a verification-based quality filter to FineWeb and Chinese FineWeb data.AI · FreeUSUltraData SFT 2605OpenBMB · FreeUltraData-SFT-2605 is OpenBMB's set of over 15 million supervised fine-tuning samples used to post-train a small on-device language model. It covers math, code, knowledge and instruction following, with both deep-thinking and non-thinking data.AI · FreeUMUltraData MathOpenBMB · FreeUltraData-Math is an OpenBMB math pre-training dataset of over 290B tokens in three quality tiers, built to improve mathematical reasoning in language models.AI · Free

Related tools

See more
X1Xperience 10MRopedia AI · FreeXperience-10M is Ropedia AI's egocentric multimodal dataset of 10 million experiences and 10,000 hours of first-person recordings. It is for work on embodied AI, robotics and world models.AI · FreePBPretrain BehaviorsIFM · FreePretrain-Behaviors is a behavior-focused text dataset from IFM covering reasoning, planning, data science, games and format rewriting. It is part of the K2 Horizon collection.AI · FreeFineWeb iconFineWebHugging Face · FreeFineWeb is Hugging Face's dataset of more than 18.5T tokens of cleaned and deduplicated English web text from Common Crawl. It is meant for pre-training language models.AI · Free