🐦 Twitter Post Details

Viewing enriched Twitter post

@f_charton

Math transformers learn better when trained from repeated examples. New paper with @KempeLab https://t.co/aTIBfmqAtJ On 3 problems, modular multiplication, GCD and eigenvalues, for the same training budget, models trained from smaller datasets achieve better performances. 1/5 https://t.co/SLZd458wcq

Media 1

📊 Media Metadata

{
  "media": [
    {
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1844349628352954678/media_0.png",
      "type": "photo",
      "original_url": "https://pbs.twimg.com/media/GZhpNKMXcAcWi7r.png",
      "recovered_from_supabase": true
    }
  ],
  "conversion_date": "2025-08-13T00:27:28.766098",
  "format_converted": true,
  "original_structure": "had_media_only"
}

🔧 Raw API Response

{
  "user": {
    "created_at": "2019-12-04T10:01:04.000Z",
    "default_profile_image": false,
    "description": "AI for mathematics and theoretical physics\nFAIR, Meta",
    "fast_followers_count": 0,
    "favourites_count": 208,
    "followers_count": 3295,
    "friends_count": 31,
    "has_custom_timelines": false,
    "is_translator": false,
    "listed_count": 57,
    "location": "Paris, France",
    "media_count": 53,
    "name": "François Charton",
    "normal_followers_count": 3295,
    "possibly_sensitive": false,
    "profile_banner_url": "https://pbs.twimg.com/profile_banners/1202165863895449600/1698860349",
    "profile_image_url_https": "https://pbs.twimg.com/profile_images/1202167628124606465/J0uFRFxL_normal.jpg",
    "screen_name": "f_charton",
    "statuses_count": 240,
    "translator_type": "none",
    "verified": false,
    "withheld_in_countries": [],
    "id_str": "1202165863895449600"
  },
  "id": "1844349628352954678",
  "conversation_id": "1844349628352954678",
  "full_text": "Math transformers learn better when trained from repeated examples. New paper with @KempeLab https://t.co/aTIBfmqAtJ\nOn 3 problems, modular multiplication, GCD and eigenvalues, for the same training budget, models trained from smaller datasets achieve better performances. 1/5 https://t.co/SLZd458wcq",
  "reply_count": 4,
  "retweet_count": 22,
  "favorite_count": 126,
  "hashtags": [],
  "symbols": [],
  "user_mentions": [
    {
      "id_str": "1782793458027036672",
      "name": "Julia Kempe",
      "screen_name": "KempeLab",
      "profile": "https://twitter.com/KempeLab"
    }
  ],
  "urls": [
    {
      "url": "https://t.co/aTIBfmqAtJ",
      "expanded_url": "https://arxiv.org/html/2410.07041v1",
      "display_url": "arxiv.org/html/2410.0704…"
    }
  ],
  "media": [
    {
      "media_url": "https://pbs.twimg.com/media/GZhpNKMXcAcWi7r.png",
      "type": "photo"
    }
  ],
  "url": "https://twitter.com/f_charton/status/1844349628352954678",
  "created_at": "2024-10-10T12:09:52.000Z",
  "#sort_index": "1844349628352954678",
  "view_count": 27906,
  "quote_count": 2,
  "is_quote_tweet": false,
  "is_retweet": false,
  "is_pinned": false,
  "is_truncated": false,
  "startUrl": "https://x.com/f_charton/status/1844349628352954678"
}