🐦 Twitter Post Details

Viewing enriched Twitter post

@dair_ai

Another fascinating paper on LLM Judges. (bookmark it) It's from Amazon, and they show that if you run panels of LLM judges, averaging their scores is a trap. "Overall, we establish that robust aggregation of a small, diverse committee is a parameter-efficient and statistically principled alternative to scaling a single large LLM-as-judge" This research proves that a mean-based panel picks up unbounded bias the moment one judge fails in a biased, LLM-typical way. Mode collapse, sycophancy, or a safety refusal from a single judge is enough, and adding more judges does not save you. RoPoLL keeps the panel but swaps the aggregator for the geometric median. It is tuning-free and hits the optimal one-half breakdown point, with a finite-sample bound and a matching minimax lower bound to back it. Across 13 judges from 4B to 675B and corruption rates up to 50 percent, RoPoLL beats the mean on every biased corruption type. A 3-judge committee at 38B outscores Mistral-Large-3 at 675B under 30 percent corruption, an 18x parameter advantage at better accuracy. Paper: https://t.co/t5QomKUoU0 Learn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c

Media 1
View on Twitter

📊 Media Metadata

{
  "media": [
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2072757239035973820/media_0.jpg",
      "filename": "media_0.jpg"
    }
  ],
  "processed_at": "2026-07-02T21:00:58.184148",
  "pipeline_version": "2.0"
}

🔧 Raw API Response

{
  "type": "tweet",
  "id": "2072757239035973820",
  "url": "https://x.com/dair_ai/status/2072757239035973820",
  "twitterUrl": "https://twitter.com/dair_ai/status/2072757239035973820",
  "text": "Another fascinating paper on LLM Judges.\n\n(bookmark it)\n\nIt's from Amazon, and they show that if you run panels of LLM judges, averaging their scores is a trap.\n\n\"Overall, we establish that robust aggregation of a small, diverse committee is a parameter-efficient and statistically principled alternative to scaling a single large LLM-as-judge\"\n\nThis research proves that a mean-based panel picks up unbounded bias the moment one judge fails in a biased, LLM-typical way. Mode collapse, sycophancy, or a safety refusal from a single judge is enough, and adding more judges does not save you.\n\nRoPoLL keeps the panel but swaps the aggregator for the geometric median. It is tuning-free and hits the optimal one-half breakdown point, with a finite-sample bound and a matching minimax lower bound to back it.\n\nAcross 13 judges from 4B to 675B and corruption rates up to 50 percent, RoPoLL beats the mean on every biased corruption type. A 3-judge committee at 38B outscores Mistral-Large-3 at 675B under 30 percent corruption, an 18x parameter advantage at better accuracy.\n\nPaper: https://t.co/t5QomKUoU0\n\nLearn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c",
  "source": "Twitter for iPhone",
  "retweetCount": 8,
  "replyCount": 2,
  "likeCount": 37,
  "quoteCount": 1,
  "viewCount": 3340,
  "createdAt": "Thu Jul 02 19:00:10 +0000 2026",
  "lang": "en",
  "bookmarkCount": 40,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2072757239035973820",
  "displayTextRange": [
    0,
    267
  ],
  "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": 127257,
    "following": 2,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Sun Jul 23 09:12:45 +0000 2017",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 4712,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 267,
    "statusesCount": 3407,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2072470330535473485"
    ],
    "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/UqTEpr8PAG",
        "expanded_url": "https://twitter.com/dair_ai/status/2072757239035973820/photo/1",
        "ext_master_playlist_only": [],
        "ext_media_availability": {
          "status": "Available"
        },
        "ext_playlists": [],
        "features": {
          "large": {
            "faces": []
          },
          "orig": {
            "faces": []
          }
        },
        "id_str": "2072757235437207552",
        "indices": [
          268,
          291
        ],
        "media_key": "3_2072757235437207552",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARzD6blD2wAACgACHMPpuhpb0LwAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHMPpuUPbAAAKAAIcw+m6GlvQvAAA",
            "media_key": "3_2072757235437207552"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HMPpuUPbAAAFarh.jpg",
        "original_info": {
          "focus_rects": [
            {
              "h": 887,
              "w": 1584,
              "x": 0,
              "y": 0
            },
            {
              "h": 1512,
              "w": 1512,
              "x": 36,
              "y": 0
            },
            {
              "h": 1512,
              "w": 1326,
              "x": 129,
              "y": 0
            },
            {
              "h": 1512,
              "w": 756,
              "x": 414,
              "y": 0
            },
            {
              "h": 1512,
              "w": 1584,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1512,
          "width": 1584
        },
        "sizes": {
          "large": {
            "h": 1512,
            "w": 1584
          }
        },
        "type": "photo",
        "url": "https://t.co/UqTEpr8PAG"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "urls": [
      {
        "display_url": "arxiv.org/abs/2606.30931",
        "expanded_url": "https://arxiv.org/abs/2606.30931",
        "indices": [
          1079,
          1102
        ],
        "url": "https://t.co/t5QomKUoU0"
      },
      {
        "display_url": "academy.dair.ai",
        "expanded_url": "https://academy.dair.ai/",
        "indices": [
          1155,
          1178
        ],
        "url": "https://t.co/LRnpZN7L4c"
      }
    ],
    "user_mentions": []
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "communityInfo": null,
  "article": null
}