{
  "type": "tweet",
  "id": "2001652694940029344",
  "url": "https://x.com/dair_ai/status/2001652694940029344",
  "twitterUrl": "https://twitter.com/dair_ai/status/2001652694940029344",
  "text": "Scaling Laws and Symmetry\n\nThe common belief is that scaling outperforms inductive biases. Give the model enough data and compute, and it will learn the structure on its own.\n\nBut this new research finds the opposite.\n\nResearchers conducted comprehensive scaling experiments on neural network interatomic potentials, comparing architectures that encode rotational and permutation symmetry to varying degrees.\n\nThe main finding is that equivariant architectures don't just have lower loss at any given scale. They have better scaling exponents. The performance gap grows as you add more compute.\n\nFor geometric tasks with known symmetries, the scaling laws favor architectures that encode those symmetries.\n\n\"Contrary to common belief, we should not leave it to the model to discover fundamental inductive biases such as symmetry, especially as we scale, because they change the inherent difficulty of the task and its scaling laws.\"\n\nPaper: https://t.co/HR5cNwFyaL\n\nLearn to build effective AI Agents in our academy: https://t.co/zQXQt0PMbG",
  "source": "Twitter for iPhone",
  "retweetCount": 1,
  "replyCount": 0,
  "likeCount": 11,
  "quoteCount": 1,
  "viewCount": 796,
  "createdAt": "Thu Dec 18 13:56:06 +0000 2025",
  "lang": "en",
  "bookmarkCount": 11,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2001652694940029344",
  "displayTextRange": [
    0,
    278
  ],
  "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/889050642903293953/1742055232",
    "description": "",
    "location": "",
    "followers": 83530,
    "following": 1,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Sun Jul 23 09:12:45 +0000 2017",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 3905,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 92,
    "statusesCount": 2676,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2001306474803540415"
    ],
    "profile_bio": {
      "description": "Democratizing AI research, education, and technologies.",
      "entities": {
        "description": {},
        "url": {
          "urls": [
            {
              "display_url": "dair.ai",
              "expanded_url": "https://www.dair.ai/",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/lkqPZtMmfU"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/MbLXt92RKK",
        "expanded_url": "https://twitter.com/dair_ai/status/2001652694940029344/photo/1",
        "ext_media_availability": {
          "status": "Available"
        },
        "features": {
          "large": {},
          "orig": {}
        },
        "id_str": "2001652691177738241",
        "indices": [
          279,
          302
        ],
        "media_key": "3_2001652691177738241",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARvHTIR1WpABCgACG8dMhVWakaAAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABG8dMhHVakAEKAAIbx0yFVZqRoAAA",
            "media_key": "3_2001652691177738241"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/G8dMhHVakAEaN6V.jpg",
        "original_info": {
          "focus_rects": [
            {
              "h": 822,
              "w": 1468,
              "x": 0,
              "y": 0
            },
            {
              "h": 1468,
              "w": 1468,
              "x": 0,
              "y": 0
            },
            {
              "h": 1674,
              "w": 1468,
              "x": 0,
              "y": 0
            },
            {
              "h": 1782,
              "w": 891,
              "x": 0,
              "y": 0
            },
            {
              "h": 1782,
              "w": 1468,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1782,
          "width": 1468
        },
        "sizes": {
          "large": {
            "h": 1782,
            "w": 1468
          }
        },
        "type": "photo",
        "url": "https://t.co/MbLXt92RKK"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "urls": [
      {
        "display_url": "openreview.net/pdf/b03ba1df90…",
        "expanded_url": "https://openreview.net/pdf/b03ba1df906f97896361a34adab7f54b7a898135.pdf",
        "indices": [
          941,
          964
        ],
        "url": "https://t.co/HR5cNwFyaL"
      },
      {
        "display_url": "dair-ai.thinkific.com",
        "expanded_url": "https://dair-ai.thinkific.com/",
        "indices": [
          1017,
          1040
        ],
        "url": "https://t.co/zQXQt0PMbG"
      }
    ]
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "article": null
}