Papers.Data.Code
Dataset

EgoSteer Real-World Bimanual Teleoperation Dataset

EgoSteer

Contains 54,454 teleoperated real-robot episodes over 193 bimanual dexterous tasks with dual egocentric RGB-D video, 74D state/action, and English instructions.

Hugging Face ↗Sep 15, 2026Paper ↗Code ↗

  • robotics
  • bimanual
  • egocentric-rgbd
  • teleoperation
Detailed pipeline diagram showing model inference, teleoperation, and human-in-the-loop correction system
Detailed pipeline diagram showing model inference, teleoperation, and human-in-the-loop correction system

What it is

Large real-robot dataset for egocentric bimanual teleoperation: 54k episodes across 193 dexterous tasks with RGB-D, proprioception, actions, and language. 54,454 real-robot teleop episodes across 193 tabletop manipulation tasks, with dual egocentric RGB-D streams, 74D state/action time series, and 1-12 English instructions per episode.

Key results

  • 54,454 episodes, 20,750,677 frames at 30 Hz
  • 193 tasks across 7 manipulation categories
  • 2 RGB-D cameras, 74D state/action, Apache-2.0

Useful for training and benchmarking language-conditioned visuomotor policies for real-world bimanual dexterous manipulation.

One useful ML email a week.

More from Sep 21 – Sep 27

The whole week →

Picked Sep 25

  • Dataset
    microsoft/ProgramDistill

    Unlike GUI-agent benchmarks, it evaluates coding agents by restoring masked web-app features with replayable browser traces and executable verifiers across 4,063 tasks.

    Hugging FaceSep 24, 2026
    →
  • Paper
    PACT: From Credit Assignment to Critic Alignment

    PACT gets 72.87% avg on 4 math benchmarks, +8.80 over GRPO and +13.16 over PPO; 67.4% on SWE-bench Verified, +2.0 over GRPO.

    arXivSep 22, 2026
    →
  • Paper
    Recursive self-improvement of AI research agents

    In 8 days it found 7 accepted rewrites, raised grade 0.703→0.778, matched/exceeded a human-built agent on 4 held-out benchmarks, and cut reward hacking 55%→32%.

    arXivSep 22, 2026
    →
  • Repo
    TencentARC/GameHorizon

    Covers 5,000 gameplay hours from 21 AAA titles, 5,000 offline MCQs, 20 online tasks, and benchmarks 47 models in one suite.

    GitHubSep 20, 2026
    →

Picked Sep 24

  • Dataset
    Harland/OmniVChat

    Makes native audio-visual dialogue measurable without text queries, ASR, or captioning, with fine-grained rubric scores shared across eval and RL.

    Hugging FaceSep 18, 2026
    →
  • Repo
    jaredpalmer/kev

    Unlike recent typed-decision repos, it packs many question types into one calibrated, decode-free pass with 2.0× speedup and 4e-6 output parity.

    GitHubSep 17, 2026
    →

Worth your time, not all of your time.

We track new ML work across papers, code, and data, then keep only the pieces worth a closer look. Selected, not collected.

Papers.Data.Code

The Weekly ML Signal

Selected, not collected.

One useful ML email a week: the papers, repos and datasets worth your attention.

Worth your time, not all of your time.