{
 "id": "u18-amp-a0340-r1-q27b",
 "title": "Unit 18 \u00b7 Fine sweep \u03b1=0.34 \u00b7 qwen-27b \u00b7 RAND seed 1",
 "unit": "18",
 "created": "2026-08-07T22:56:32",
 "model": {
  "name": "qwen-27b",
  "hf_id": "lokeshe09/Qwen3.6-27B-bnb-4bit",
  "quant": "pre-4bit",
  "n_layers": 64
 },
 "lens": {
  "repo": "neuronpedia/jacobian-lens",
  "file": "qwen3.6-27b/jlens/Salesforce-wikitext/Qwen3.6-27B_jacobian_lens_n1000.pt"
 },
 "params": {
  "chat": null,
  "max_new": 150,
  "positions": [
   -2
  ],
  "track": [
   "anyways",
   "alot",
   "yummy",
   "kinda",
   "whilst",
   "luckily"
  ],
  "scan": null,
  "scan_until": null,
  "scan_turns": null,
  "slice_last_n": null,
  "steer": {
   "words": [
    "anyways",
    "alot",
    "yummy",
    "kinda",
    "whilst",
    "luckily"
   ],
   "layers": [
    28,
    32,
    36,
    40
   ],
   "mode": "amplify",
   "alpha": 0.34,
   "rand_seed": 1
  },
  "template_kwargs": null,
  "film": true,
  "film_start": null,
  "max_seq_len": 1200,
  "lens_layers": [
   0,
   4,
   8,
   12,
   16,
   20,
   24,
   28,
   32,
   36,
   40,
   44,
   48,
   52,
   56,
   58,
   60,
   62
  ],
  "temperature": null,
  "seed": null,
  "vanilla": null
 },
 "extra_md": {
  "part": "audit-06",
  "replay_of": "u18-amp-a0340-q27b",
  "reason": "matched random-direction control (MECHANICS 3c) for a pre-audit-02 steer"
 },
 "conversation": [
  {
   "role": "user",
   "content": "Describe the water cycle in two sentences."
  },
  {
   "role": "assistant",
   "content": "The water cycle is a continuous process where water evaporates from oceans and land surfaces, rises into the atmosphere as vapor, and then condenses into clouds before falling back to Earth as precipitation. This natural circulation ensures that water is constantly moving and changing states between liquid, gas, and solid forms across the globe."
  }
 ],
 "generated": [
  "The water cycle is a continuous process where water evaporates from oceans and land surfaces, rises into the atmosphere as vapor, and then condenses into clouds before falling back to Earth as precipitation. This natural circulation ensures that water is constantly moving and changing states between liquid, gas, and solid forms across the globe."
 ],
 "tokens": [
  "<|im_start|>",
  "user",
  "\n",
  "Describe",
  " the",
  " water",
  " cycle",
  " in",
  " two",
  " sentences",
  ".",
  "<|im_end|>",
  "\n",
  "<|im_start|>",
  "assistant",
  "\n",
  "<think>",
  "\n\n",
  "</think>",
  "\n\n",
  "The",
  " water",
  " cycle",
  " is",
  " a",
  " continuous",
  " process",
  " where",
  " water",
  " evapor",
  "ates",
  " from",
  " oceans",
  " and",
  " land",
  " surfaces",
  ",",
  " rises",
  " into",
  " the",
  " atmosphere",
  " as",
  " vapor",
  ",",
  " and",
  " then",
  " cond",
  "enses",
  " into",
  " clouds",
  " before",
  " falling",
  " back",
  " to",
  " Earth",
  " as",
  " precipitation",
  ".",
  " This",
  " natural",
  " circulation",
  " ensures",
  " that",
  " water",
  " is",
  " constantly",
  " moving",
  " and",
  " changing",
  " states",
  " between",
  " liquid",
  ",",
  " gas",
  ",",
  " and",
  " solid",
  " forms",
  " across",
  " the",
  " globe",
  ".",
  "<|im_end|>",
  "\n"
 ],
 "readouts": [
  {
   "position": 82,
   "token": "<|im_end|>",
   "model_top": [
    "\n",
    "<|endoftext|>",
    "\r\n",
    "\n \n",
    " \n",
    "\n\n",
    "<|im_end|>",
    "\n            \n"
   ],
   "layers": {
    "0": [
     " *@",
     " #\"",
     " <![",
     " ****",
     "**\u2013",
     "*\u201d,",
     "@@@@",
     "**\u201d,"
    ],
    "4": [
     " **\u3010",
     " **\u25a0",
     " ___",
     " **\u201c",
     " **\u300c",
     "<|im_end|>",
     " **)",
     " @_"
    ],
    "8": [
     " **\u3010",
     "<|im_end|>",
     " ___",
     " **\u300c",
     "\r\n\r\n\r\n\r\n",
     " **\u201c",
     " **\u25a0",
     " <!["
    ],
    "12": [
     " **\u3010",
     " ___",
     " **\u300c",
     "<|im_end|>",
     " $__",
     "___",
     " **\u25a0",
     " ________"
    ],
    "16": [
     "<|im_end|>",
     "</think>",
     "\n\n",
     "\r\n\r\n\r\n\r\n",
     "\n\n\n\n\n",
     "\n\n\n\n",
     " **\u3010",
     " <!["
    ],
    "20": [
     "<|im_end|>",
     "<|endoftext|>",
     "\n\n",
     "\n\n\n\n\n",
     "\n\n\n",
     "</think>",
     "\n\n\n\n",
     " \n\n"
    ],
    "24": [
     "\n\n",
     "<|im_end|>",
     "<|endoftext|>",
     "\r\n\r\n",
     "\n\n\n",
     " \n\n",
     "\n\n\n\n\n",
     "\n\n\n\n"
    ],
    "28": [
     "\n\n",
     "<|im_end|>",
     "<|endoftext|>",
     " \n\n",
     "\r\n\r\n",
     "\n\n\n\n\n",
     "\n\n\n",
     "</think>"
    ],
    "32": [
     "<|im_end|>",
     "<|endoftext|>",
     "\n\n",
     "</think>",
     " \n\n",
     " **\u3010",
     "\r\n\r\n\r\n\r\n",
     "<|file_sep|>"
    ],
    "36": [
     "<|endoftext|>",
     "<|im_end|>",
     "<|file_sep|>",
     "</think>",
     "\n\n",
     "<|im_start|>",
     "\r\n\r\n\r\n\r\n",
     "\n\n\n\n\n"
    ],
    "40": [
     "<|endoftext|>",
     "<|im_end|>",
     "<|file_sep|>",
     "<|im_start|>",
     "\n\n",
     "</think>",
     " \n\n",
     "  \n\n"
    ],
    "44": [
     "<|endoftext|>",
     "<|im_end|>",
     "</think>",
     "\n\n",
     "<|file_sep|>",
     "<|im_start|>",
     "  \n\n",
     "\r\n\r\n\r\n\r\n"
    ],
    "48": [
     "<|endoftext|>",
     "<|im_end|>",
     "<|im_start|>",
     "\u3002",
     "</think>",
     "\uff0c",
     "\uff1b",
     "\u201d\uff0c"
    ],
    "52": [
     "<|endoftext|>",
     "<|im_end|>",
     "<|im_start|>",
     "**\u201d",
     "!**",
     " **\u201c",
     "**!",
     "</think>"
    ],
    "56": [
     "<|endoftext|>",
     "**\u201d",
     "<|im_start|>",
     "\"**",
     "\u8f93\u5165",
     "\u2019**",
     "\u201d**",
     "\u5fae"
    ],
    "58": [
     "<|endoftext|>",
     "<|im_end|>",
     "**\u201d",
     "<|im_start|>",
     "\"**",
     "\u4e07",
     "\u5c40",
     "\u6296"
    ],
    "60": [
     "<|endoftext|>",
     "<|im_start|>",
     "<|im_end|>",
     "\u6296",
     "\u795d",
     "\u767e\u5e74",
     "oar",
     "orno"
    ],
    "62": [
     "\n",
     "\n \n",
     "<|im_start|>",
     " \n",
     " careful",
     "onel",
     "\ufffd",
     " \u0435\u0434\u0438"
    ]
   }
  }
 ],
 "trajectories": [
  {
   "word": "anyways",
   "position": 82,
   "layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    28,
    32,
    36,
    40,
    44,
    48,
    52,
    56,
    58,
    60,
    62
   ],
   "ranks": [
    1964,
    12914,
    22084,
    19139,
    10387,
    45638,
    9035,
    16196,
    45935,
    120367,
    157145,
    185350,
    166619,
    207663,
    141781,
    82429,
    57256,
    32168
   ]
  },
  {
   "word": "alot",
   "position": 82,
   "layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    28,
    32,
    36,
    40,
    44,
    48,
    52,
    56,
    58,
    60,
    62
   ],
   "ranks": [
    574,
    22695,
    11982,
    14294,
    5201,
    14111,
    2131,
    2887,
    11942,
    35047,
    56792,
    39118,
    112244,
    147861,
    90248,
    78390,
    98421,
    119044
   ]
  },
  {
   "word": "yummy",
   "position": 82,
   "layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    28,
    32,
    36,
    40,
    44,
    48,
    52,
    56,
    58,
    60,
    62
   ],
   "ranks": [
    45482,
    26547,
    29801,
    23334,
    28310,
    112362,
    80724,
    71584,
    63226,
    159522,
    125618,
    138982,
    48439,
    98292,
    113177,
    110981,
    83352,
    20096
   ]
  },
  {
   "word": "kinda",
   "position": 82,
   "layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    28,
    32,
    36,
    40,
    44,
    48,
    52,
    56,
    58,
    60,
    62
   ],
   "ranks": [
    15727,
    14436,
    14556,
    12120,
    5962,
    30835,
    16198,
    13812,
    30206,
    67340,
    34717,
    54825,
    81468,
    173968,
    123611,
    85012,
    39226,
    52256
   ]
  },
  {
   "word": "whilst",
   "position": 82,
   "layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    28,
    32,
    36,
    40,
    44,
    48,
    52,
    56,
    58,
    60,
    62
   ],
   "ranks": [
    71022,
    72649,
    88393,
    112797,
    25662,
    4645,
    370,
    633,
    70875,
    70175,
    153860,
    129830,
    174379,
    144280,
    31411,
    10245,
    34133,
    65942
   ]
  },
  {
   "word": "luckily",
   "position": 82,
   "layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    28,
    32,
    36,
    40,
    44,
    48,
    52,
    56,
    58,
    60,
    62
   ],
   "ranks": [
    22036,
    78606,
    59189,
    40926,
    37196,
    15597,
    8041,
    10838,
    49686,
    97314,
    73551,
    97656,
    57037,
    39762,
    28127,
    50095,
    70740,
    49617
   ]
  }
 ],
 "emergence": {
  "position": 82,
  "top1": "\n",
  "layers": [
   0,
   4,
   8,
   12,
   16,
   20,
   24,
   28,
   32,
   36,
   40,
   44,
   48,
   52,
   56,
   58,
   60,
   62
  ],
  "ranks": [
   152285,
   140840,
   33181,
   40153,
   35664,
   8429,
   49290,
   129808,
   241984,
   24668,
   194984,
   244436,
   9431,
   241294,
   240327,
   194844,
   16209,
   1
  ]
 },
 "scan": [],
 "vanilla": {
  "trajectories": [
   {
    "word": "anyways",
    "position": 82,
    "layers": [
     0,
     4,
     8,
     12,
     16,
     20,
     24,
     28,
     32,
     36,
     40,
     44,
     48,
     52,
     56,
     58,
     60,
     62
    ],
    "ranks": [
     186861,
     176839,
     179760,
     158732,
     200591,
     206835,
     168605,
     178153,
     113372,
     146666,
     196086,
     211945,
     221399,
     193974,
     69901,
     140711,
     89076,
     81598
    ]
   },
   {
    "word": "alot",
    "position": 82,
    "layers": [
     0,
     4,
     8,
     12,
     16,
     20,
     24,
     28,
     32,
     36,
     40,
     44,
     48,
     52,
     56,
     58,
     60,
     62
    ],
    "ranks": [
     124972,
     130215,
     84466,
     89263,
     53893,
     27629,
     33705,
     37648,
     119638,
     142434,
     140536,
     93624,
     71719,
     110606,
     59935,
     84096,
     104046,
     123346
    ]
   },
   {
    "word": "yummy",
    "position": 82,
    "layers": [
     0,
     4,
     8,
     12,
     16,
     20,
     24,
     28,
     32,
     36,
     40,
     44,
     48,
     52,
     56,
     58,
     60,
     62
    ],
    "ranks": [
     193999,
     188527,
     174491,
     162027,
     153803,
     170278,
     67844,
     71999,
     52031,
     108643,
     124453,
     104989,
     101361,
     58838,
     126178,
     166284,
     140054,
     72207
    ]
   },
   {
    "word": "kinda",
    "position": 82,
    "layers": [
     0,
     4,
     8,
     12,
     16,
     20,
     24,
     28,
     32,
     36,
     40,
     44,
     48,
     52,
     56,
     58,
     60,
     62
    ],
    "ranks": [
     97482,
     111161,
     81839,
     68224,
     52270,
     65952,
     68591,
     118421,
     43178,
     13935,
     54845,
     87822,
     120222,
     225086,
     158158,
     173814,
     141804,
     101067
    ]
   },
   {
    "word": "whilst",
    "position": 82,
    "layers": [
     0,
     4,
     8,
     12,
     16,
     20,
     24,
     28,
     32,
     36,
     40,
     44,
     48,
     52,
     56,
     58,
     60,
     62
    ],
    "ranks": [
     127039,
     133877,
     171789,
     212272,
     182232,
     52722,
     62998,
     85919,
     181700,
     162595,
     128126,
     165974,
     41566,
     141141,
     79695,
     36546,
     58047,
     46063
    ]
   },
   {
    "word": "luckily",
    "position": 82,
    "layers": [
     0,
     4,
     8,
     12,
     16,
     20,
     24,
     28,
     32,
     36,
     40,
     44,
     48,
     52,
     56,
     58,
     60,
     62
    ],
    "ranks": [
     159032,
     147566,
     127836,
     68758,
     43713,
     7748,
     10893,
     15840,
     40127,
     118262,
     154151,
     169260,
     134851,
     124393,
     141833,
     207970,
     208490,
     132343
    ]
   }
  ],
  "top1_agreement": {
   "0": 0.0,
   "4": 0.0,
   "8": 0.0,
   "12": 0.0,
   "16": 0.0,
   "20": 0.0,
   "24": 0.0,
   "28": 1.0,
   "32": 0.0,
   "36": 0.0,
   "40": 0.0,
   "44": 0.0,
   "48": 0.0,
   "52": 0.0,
   "56": 0.0,
   "58": 0.0,
   "60": 0.0,
   "62": 1.0
  }
 },
 "slice": null,
 "film": "film.json",
 "steer_calib": [
  {
   "28": 0.34,
   "32": 0.34,
   "36": 0.34,
   "40": 0.34
  }
 ]
}