{
  "type": "tweet",
  "id": "2052435456634110336",
  "url": "https://x.com/AnthropicAI/status/2052435456634110336",
  "twitterUrl": "https://twitter.com/AnthropicAI/status/2052435456634110336",
  "text": "NLA training doesn’t guarantee that explanations are faithful descriptions of Claude’s thoughts. But based on experience and experimental evidence, we think they often are.\n\nFor instance, we find that NLAs help discover hidden motivations in an intentionally misaligned model. https://t.co/NJ5yc8p7Dn",
  "source": "Twitter for iPhone",
  "retweetCount": 8,
  "replyCount": 9,
  "likeCount": 419,
  "quoteCount": 4,
  "viewCount": 59086,
  "createdAt": "Thu May 07 17:08:39 +0000 2026",
  "lang": "en",
  "bookmarkCount": 31,
  "isReply": true,
  "inReplyToId": "2052435452657795246",
  "conversationId": "2052435436157452769",
  "displayTextRange": [
    0,
    276
  ],
  "inReplyToUserId": "1353836358901501952",
  "inReplyToUsername": "AnthropicAI",
  "author": {
    "type": "user",
    "userName": "AnthropicAI",
    "url": "https://x.com/AnthropicAI",
    "twitterUrl": "https://twitter.com/AnthropicAI",
    "id": "1353836358901501952",
    "name": "Anthropic",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": "Business",
    "profilePicture": "https://pbs.twimg.com/profile_images/1798110641414443008/XP8gyBaY_normal.jpg",
    "coverPicture": "https://pbs.twimg.com/profile_banners/1353836358901501952/1719228429",
    "description": "",
    "location": "",
    "followers": 1251930,
    "following": 36,
    "status": "",
    "canDm": false,
    "canMediaTag": true,
    "createdAt": "Mon Jan 25 22:45:28 +0000 2021",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 1573,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 621,
    "statusesCount": 1556,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [],
    "profile_bio": {
      "description": "We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems. Talk to our AI assistant @claudeai on https://t.co/FhDI3KQh0n.",
      "entities": {
        "description": {
          "urls": [
            {
              "display_url": "claude.ai",
              "expanded_url": "https://claude.ai",
              "indices": [
                141,
                164
              ],
              "url": "https://t.co/FhDI3KQh0n"
            }
          ],
          "user_mentions": [
            {
              "id_str": "",
              "indices": [
                128,
                137
              ],
              "name": "",
              "screen_name": "claudeai"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "anthropic.com",
              "expanded_url": "https://anthropic.com",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/w94SABjAXZ"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "allow_download_status": {
          "allow_download": true
        },
        "display_url": "pic.twitter.com/NJ5yc8p7Dn",
        "expanded_url": "https://twitter.com/AnthropicAI/status/2052435456634110336/photo/1",
        "ext_alt_text": "Auditors with NLAs can discover the target model’s root cause of misalignment, even without access to training data. (NLAs and SAEs for both conditions were trained using generic data only, not data used to train the target model.)",
        "ext_media_availability": {
          "status": "Available"
        },
        "features": {
          "large": {
            "faces": []
          },
          "orig": {
            "faces": []
          }
        },
        "id_str": "2052432130773331972",
        "indices": [
          277,
          300
        ],
        "media_key": "3_2052432130773331972",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARx7tCYQV7AECgACHHu3LG0XwYAAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHHu0JhBXsAQKAAIce7csbRfBgAAA",
            "media_key": "3_2052432130773331972"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HHu0JhBXsAQbzAz.jpg",
        "original_info": {
          "focus_rects": [
            {
              "h": 890,
              "w": 1590,
              "x": 0,
              "y": 0
            },
            {
              "h": 1170,
              "w": 1170,
              "x": 0,
              "y": 0
            },
            {
              "h": 1170,
              "w": 1026,
              "x": 3,
              "y": 0
            },
            {
              "h": 1170,
              "w": 585,
              "x": 224,
              "y": 0
            },
            {
              "h": 1170,
              "w": 1590,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1170,
          "width": 1590
        },
        "sizes": {
          "large": {
            "h": 1170,
            "w": 1590
          }
        },
        "type": "photo",
        "url": "https://t.co/NJ5yc8p7Dn"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "timestamps": [],
    "urls": [],
    "user_mentions": []
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "communityInfo": null,
  "article": null
}