🐦 Twitter Post Details

Viewing enriched Twitter post

@dair_ai

// Self-play with a pinch of human data // Really cool paper combining human demonstrations and self-play RL. 30 minutes of human data, 2500x less than imitation learning, is enough to make self-play policies coordinate with real people. Pure self-play learns effective but alien conventions that humans cannot drive alongside. The usual fix is brittle reward engineering and domain randomization. This work instead treats a small set of human demonstrations as a regularization objective on top of a minimal safe goal-reaching reward. Why does it matter? The resulting policies coordinate with held-out human trajectories and finish training in 15 hours on a single consumer GPU. The lesson travels well past driving. A small demonstration regularizer may be the cheapest alignment knob we have for self-play. Paper: https://t.co/nLrVwRFEW9 Learn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c

Media 1

📊 Media Metadata

{
  "media": [
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2068456364691935326/media_0.jpg",
      "filename": "media_0.jpg"
    }
  ],
  "processed_at": "2026-06-20T22:16:57.647060",
  "pipeline_version": "2.0"
}

🔧 Raw API Response

{
  "type": "tweet",
  "id": "2068456364691935326",
  "url": "https://x.com/dair_ai/status/2068456364691935326",
  "twitterUrl": "https://twitter.com/dair_ai/status/2068456364691935326",
  "text": "// Self-play with a pinch of human data //\n\nReally cool paper combining human demonstrations and self-play RL.\n\n30 minutes of human data, 2500x less than imitation learning, is enough to make self-play policies coordinate with real people.\n\nPure self-play learns effective but alien conventions that humans cannot drive alongside. The usual fix is brittle reward engineering and domain randomization.\n\nThis work instead treats a small set of human demonstrations as a regularization objective on top of a minimal safe goal-reaching reward.\n\nWhy does it matter?\n\nThe resulting policies coordinate with held-out human trajectories and finish training in 15 hours on a single consumer GPU. The lesson travels well past driving. A small demonstration regularizer may be the cheapest alignment knob we have for self-play.\n\nPaper: https://t.co/nLrVwRFEW9\n\nLearn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c",
  "source": "Twitter for iPhone",
  "retweetCount": 0,
  "replyCount": 0,
  "likeCount": 1,
  "quoteCount": 0,
  "viewCount": 191,
  "createdAt": "Sat Jun 20 22:10:02 +0000 2026",
  "lang": "en",
  "bookmarkCount": 4,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2068456364691935326",
  "displayTextRange": [
    0,
    276
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "dair_ai",
    "url": "https://x.com/dair_ai",
    "twitterUrl": "https://twitter.com/dair_ai",
    "id": "889050642903293953",
    "name": "DAIR.AI",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/1643277398522187778/31dedbLo_normal.jpg",
    "coverPicture": "https://pbs.twimg.com/profile_banners/889050642903293953/1773242460",
    "description": "",
    "location": "",
    "followers": 126361,
    "following": 1,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Sun Jul 23 09:12:45 +0000 2017",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 4622,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 252,
    "statusesCount": 3341,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2067984002376749525"
    ],
    "profile_bio": {
      "description": "Democratizing AI research, education, and technologies. Learn about AI Agents for FREE at https://t.co/HHXg8rryu4",
      "entities": {
        "description": {
          "urls": [
            {
              "display_url": "academy.dair.ai/courses/elemen…",
              "expanded_url": "https://academy.dair.ai/courses/elements-of-ai-agents",
              "indices": [
                90,
                113
              ],
              "url": "https://t.co/HHXg8rryu4"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "dair.ai",
              "expanded_url": "https://www.dair.ai/",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/lkqPZtMU5s"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/k83mWajt26",
        "expanded_url": "https://twitter.com/dair_ai/status/2068456364691935326/photo/1",
        "ext_master_playlist_only": [],
        "ext_media_availability": {
          "status": "Available"
        },
        "ext_playlists": [],
        "features": {
          "large": {
            "faces": [
              {
                "h": 219,
                "w": 219,
                "x": 202,
                "y": 1042
              }
            ]
          },
          "orig": {
            "faces": [
              {
                "h": 219,
                "w": 219,
                "x": 202,
                "y": 1042
              }
            ]
          }
        },
        "id_str": "2068456361097330688",
        "indices": [
          277,
          300
        ],
        "media_key": "3_2068456361097330688",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARy0ohnsmoAACgACHLSiGsLb0F4AAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHLSiGeyagAAKAAIctKIawtvQXgAA",
            "media_key": "3_2068456361097330688"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HLSiGeyagAANUVI.jpg",
        "original_info": {
          "focus_rects": [
            {
              "h": 905,
              "w": 1616,
              "x": 0,
              "y": 0
            },
            {
              "h": 1616,
              "w": 1616,
              "x": 0,
              "y": 0
            },
            {
              "h": 1842,
              "w": 1616,
              "x": 0,
              "y": 0
            },
            {
              "h": 1870,
              "w": 935,
              "x": 139,
              "y": 0
            },
            {
              "h": 1870,
              "w": 1616,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1870,
          "width": 1616
        },
        "sizes": {
          "large": {
            "h": 1870,
            "w": 1616
          }
        },
        "type": "photo",
        "url": "https://t.co/k83mWajt26"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "urls": [
      {
        "display_url": "arxiv.org/abs/2606.19370",
        "expanded_url": "https://arxiv.org/abs/2606.19370",
        "indices": [
          825,
          848
        ],
        "url": "https://t.co/nLrVwRFEW9"
      },
      {
        "display_url": "academy.dair.ai",
        "expanded_url": "https://academy.dair.ai/",
        "indices": [
          901,
          924
        ],
        "url": "https://t.co/LRnpZN7L4c"
      }
    ],
    "user_mentions": []
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "communityInfo": null,
  "article": null
}