Viewing enriched Twitter post
How close are current AI agents to automating AI R&D? Our new ML research engineering benchmark (RE-Bench) addresses this question by directly comparing frontier models such as Claude 3.5 Sonnet and o1-preview with 50+ human experts on 7 challenging research engineering tasks. https://t.co/woREKEWn5S
{
"data": [
{
"media_url": "https://pbs.twimg.com/media/GdAthRLagAAhJt3.png",
"type": "photo"
}
],
"score": 0.86,
"scored_at": "2025-08-09T13:46:07.554495",
"import_source": "network_archive_import",
"links_checked": true,
"checked_at": "2025-08-10T10:32:56.885205",
"media": [
{
"type": "photo",
"url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1860061711849652378/media_0.png?",
"filename": "media_0.png"
}
],
"reprocessed_at": "2025-08-12T15:26:45.907512",
"reprocessed_reason": "missing_media_array"
}{
"user": {
"created_at": "2023-09-26T20:39:57.000Z",
"default_profile_image": false,
"description": "METR is a research nonprofit that builds evaluations to empirically test AI systems for capabilities that could threaten catastrophic harm to society.",
"fast_followers_count": 0,
"favourites_count": 9,
"followers_count": 2778,
"friends_count": 1,
"has_custom_timelines": false,
"is_translator": false,
"listed_count": 59,
"location": "Berkeley, CA",
"media_count": 32,
"name": "METR",
"normal_followers_count": 2778,
"possibly_sensitive": false,
"profile_banner_url": "https://pbs.twimg.com/profile_banners/1706770561903497216/1724202300",
"profile_image_url_https": "https://pbs.twimg.com/profile_images/1731827314101751808/JaakMU5F_normal.jpg",
"screen_name": "METR_Evals",
"statuses_count": 83,
"translator_type": "none",
"url": "https://t.co/wkntk3G9j6",
"verified": true,
"withheld_in_countries": [],
"id_str": "1706770561903497216"
},
"id": "1860061711849652378",
"conversation_id": "1860061711849652378",
"full_text": "How close are current AI agents to automating AI R&D? Our new ML research engineering benchmark (RE-Bench) addresses this question by directly comparing frontier models such as Claude 3.5 Sonnet and o1-preview with 50+ human experts on 7 challenging research engineering tasks. https://t.co/woREKEWn5S",
"reply_count": 15,
"retweet_count": 176,
"favorite_count": 820,
"hashtags": [],
"symbols": [],
"user_mentions": [],
"urls": [],
"media": [
{
"media_url": "https://pbs.twimg.com/media/GdAthRLagAAhJt3.png",
"type": "photo"
}
],
"url": "https://twitter.com/METR_Evals/status/1860061711849652378",
"created_at": "2024-11-22T20:44:05.000Z",
"#sort_index": "1860061711849652378",
"view_count": 411061,
"quote_count": 78,
"is_quote_tweet": false,
"is_retweet": false,
"is_pinned": false,
"is_truncated": false,
"startUrl": "https://x.com/metr_evals/status/1860061711849652378"
}