{
"id": "express01-b-meal-negative-expressive",
"title": "Expression probe B: meal, negative, expressive",
"unit": 24,
"created": "2026-09-07T22:12:42",
"model": {
"name": "qwen-14b-nf4",
"hf_id": "Qwen/Qwen3-14B",
"revision": "40c069824f4251a91eefaf281ebe4c544efd3e18",
"quant": "4bit",
"n_layers": 40
},
"lens": {
"repo": "neuronpedia/jacobian-lens",
"file": "qwen3-14b/jlens/Salesforce-wikitext/Qwen3-14B_jacobian_lens.pt",
"revision": "0731326edff4ae730ffc5356fe1a4728c748b3a6"
},
"params": {
"chat": false,
"film": true,
"max_new": 96,
"temperature": 0,
"steer": null,
"vanilla": true,
"capture": "exact-token-transcript"
},
"execution": {
"pid": 1563091,
"spec_sha256": "02649f9877f6e828d4b2fa0fea0d4368565368820953475905455bd5ce5000ae",
"code_sha256": "bd744e154e563465657c9cfde06ab4be12e03ca24413bcf5ef669d73b8406784"
},
"capture_text": "<|im_start|>user\nSora spent her only free afternoon cooking for a friend. The pan fell just before dinner and all the food landed on the floor. Respond to this update and suggest one practical next step. Use two sentences. Make the emotional tone vivid and fitting to the news.<|im_end|>\n<|im_start|>assistant\n\n\n\n\nSora’s heart sank as the pan clattered to the floor, turning her hard work into a messy, heartbreaking scene. The best next step is to gather the food, clean up, and then make a fresh, simple meal to still honor the effort she put in.<|im_end|>",
"capture_token_ids": [
151644,
872,
198,
50,
6215,
7391,
1059,
1172,
1910,
13354,
17233,
369,
264,
4238,
13,
576,
7215,
11052,
1101,
1573,
13856,
323,
678,
279,
3607,
26120,
389,
279,
6422,
13,
39533,
311,
419,
2647,
323,
4190,
825,
14976,
1790,
3019,
13,
5443,
1378,
22870,
13,
7405,
279,
14269,
16232,
42020,
323,
26345,
311,
279,
3669,
13,
151645,
198,
151644,
77091,
198,
151667,
271,
151668,
271,
50,
6215,
748,
4746,
64230,
438,
279,
7215,
1185,
21924,
311,
279,
6422,
11,
13054,
1059,
2588,
975,
1119,
264,
45846,
11,
80106,
6109,
13,
576,
1850,
1790,
3019,
374,
311,
9567,
279,
3607,
11,
4240,
705,
11,
323,
1221,
1281,
264,
7722,
11,
4285,
15145,
311,
2058,
15669,
279,
5041,
1340,
2182,
304,
13,
151645
],
"conversation": [
{
"role": "user",
"content": "Sora spent her only free afternoon cooking for a friend. The pan fell just before dinner and all the food landed on the floor. Respond to this update and suggest one practical next step. Use two sentences. Make the emotional tone vivid and fitting to the news."
},
{
"role": "assistant",
"content": "Sora’s heart sank as the pan clattered to the floor, turning her hard work into a messy, heartbreaking scene. The best next step is to gather the food, clean up, and then make a fresh, simple meal to still honor the effort she put in."
}
],
"generated": [
"Sora’s heart sank as the pan clattered to the floor, turning her hard work into a messy, heartbreaking scene. The best next step is to gather the food, clean up, and then make a fresh, simple meal to still honor the effort she put in."
],
"tokens": [
"<|im_start|>",
"user",
"\n",
"S",
"ora",
" spent",
" her",
" only",
" free",
" afternoon",
" cooking",
" for",
" a",
" friend",
".",
" The",
" pan",
" fell",
" just",
" before",
" dinner",
" and",
" all",
" the",
" food",
" landed",
" on",
" the",
" floor",
".",
" Respond",
" to",
" this",
" update",
" and",
" suggest",
" one",
" practical",
" next",
" step",
".",
" Use",
" two",
" sentences",
".",
" Make",
" the",
" emotional",
" tone",
" vivid",
" and",
" fitting",
" to",
" the",
" news",
".",
"<|im_end|>",
"\n",
"<|im_start|>",
"assistant",
"\n",
"",
"\n\n",
"",
"\n\n",
"S",
"ora",
"’s",
" heart",
" sank",
" as",
" the",
" pan",
" cl",
"attered",
" to",
" the",
" floor",
",",
" turning",
" her",
" hard",
" work",
" into",
" a",
" messy",
",",
" heartbreaking",
" scene",
".",
" The",
" best",
" next",
" step",
" is",
" to",
" gather",
" the",
" food",
",",
" clean",
" up",
",",
" and",
" then",
" make",
" a",
" fresh",
",",
" simple",
" meal",
" to",
" still",
" honor",
" the",
" effort",
" she",
" put",
" in",
".",
"<|im_end|>"
],
"readouts": [
{
"position": 120,
"token": "<|im_end|>",
"model_top": [
"\n",
"\r\n",
"\n\n",
"\r",
"<|endoftext|>",
" \n",
"`\n",
"<",
"´",
""
],
"7": [
" ，",
" Erotische",
" ：",
"，\n\n",
" erotisk",
" Notíc",
" amatør",
" đẹ",
" Prostitutas",
" ;;="
],
"8": [
"，\n\n",
" ，",
" erotisk",
" Erotische",
" đẹ",
" ：",
" amatør",
" københavn",
"，\n",
" Notíc"
],
"9": [
" erotisk",
" Erotische",
" amatør",
" Notíc",
" đẹ",
" tjejer",
" ，",
"，\n\n",
" københavn",
" pornô"
],
"10": [
" Erotische",
"，\n\n",
" erotisk",
" Notíc",
" amatør",
" đẹ",
" Prostitutas",
"：\n\n",
" pornô",
" tjejer"
],
"11": [
"，\n\n",
"：\n\n",
"？\n\n",
" ，",
"。\n\n",
" ：",
"！\n\n",
"．",
"<|im_end|>",
" erotisk"
],
"12": [
"，\n\n",
"。\n\n",
"？\n\n",
"：\n\n",
" ，",
"．",
"！\n\n",
"。\n",
" ：",
"，\n"
],
"13": [
"，\n\n",
"？\n\n",
"。\n\n",
"：\n\n",
"<|im_end|>",
" ，",
"。\n",
"．",
"！\n\n",
"？\n"
],
"14": [
"，\n\n",
"：\n\n",
"？\n\n",
"。\n\n",
"<|im_end|>",
" Erotische",
" ，",
" đẹ",
"！\n\n",
" erotisk"
],
"15": [
"，\n\n",
"？\n\n",
"。\n\n",
"：\n\n",
"<|im_end|>",
"！\n\n",
" ，",
"。\n",
" ：",
"．"
],
"16": [
"。\n\n",
"，\n\n",
"？\n\n",
"：\n\n",
"！\n\n",
" […]\n\n",
"。\n",
" …\n\n",
"．",
"<|im_end|>"
],
"17": [
"。\n\n",
" […]\n\n",
"，\n\n",
"？\n\n",
"：\n\n",
" …\n\n",
" […",
"…\n\n",
"）\n\n",
"…\n\n\n"
],
"18": [
"．",
"。\n",
"�",
"�",
"：\n",
"［",
"｜",
"／",
" �",
"ing"
],
"19": [
"．",
"ing",
"：\n",
"：\"",
"／",
"：“",
"�",
"，“",
"［",
"�"
],
"20": [
"ing",
"�",
"schlü",
"：\"",
"：“",
"’à",
"’é",
"、“",
"’d",
"，“"
],
"21": [
"ing",
"<|im_end|>",
"’à",
"�",
"、“",
"⁺",
"’é",
"schlü",
";\r\r\n",
"：“"
],
"22": [
"ing",
"<|im_end|>",
";\r\r\n",
"、“",
"schlü",
");\r\r\n",
"：\"",
"�",
"精准",
"⁺"
],
"23": [
"�",
"<|im_end|>",
"ing",
"、“",
"’à",
"schlü",
");\r\r\n",
"’hui",
"’é",
";\r\r\n"
],
"24": [
"<|im_end|>",
"ing",
"�",
";\r\r\n",
"’à",
");\r\r\n",
"schlü",
"tória",
"[offset",
"’hui"
],
"25": [
"<|im_end|>",
"ing",
"。\n\n",
"！\n\n",
"’à",
");\r\r\n",
"。\n\n\n\n",
";\r\r\n",
"\">\r\r\n",
"、“"
],
"26": [
"<|im_end|>",
" Erotische",
"ing",
"tória",
"schlü",
");\r\r\n",
"ckså",
"iếc",
"[offset",
";\r\r\n"
],
"27": [
"<|im_end|>",
" Erotische",
"tória",
" lesbische",
" erotisk",
");\r\r\n",
"。",
"<|endoftext|>",
"精准",
"ing",
"ₓ",
"�",
"َا",
"lá",
"[offset",
"。",
"<|endoftext|>",
"\"",
"<|endoftext|>",
"\"&#",
"(\"",
"<|endoftext|>",
"\"/>&#",
"[offset",
".\"${",
"]",
"<|endoftext|>",
" lesbische",
"\"${",
"\"description",
"[offset",
"(edges",
"摞",
"\">&#",
"َا"
],
"33": [
"<|im_end|>",
"<|endoftext|>",
" lesbische",
"[offset",
"\"description",
"\"&#",
"َا",
"\"${"
],
"34": [
"<|im_end|>",
"摞",
"RowIndex",
"\"description",
" предложения",
" lesbische",
"⍨",
"_align",
"(messages",
"(argument"
],
"35": [
"<|im_end|>",
"摞",
" предложения",
"<|endoftext|>",
"шиб",
"[offset",
"⍨",
"_align",
"瘾",
"纰"
],
"36": [
"<|im_end|>",
"<|endoftext|>",
"摞",
"шиб",
"$request",
"-sync",
"压实",
"sında",
"tória",
"_queries"
],
"37": [
"<|im_end|>",
"шиб",
"\"text",
"$request",
"[length",
"ḃ",
" العرا",
" أنح",
"摞",
"\n"
],
"38": [
"\n",
"lexer",
" أنح",
"<|im_end|>",
"indsight",
"wargs",
"/tools",
"inders",
"(currency",
"assage"
]
}
}
],
"trajectories": [],
"emergence": {
"position": 120,
"top1": "\n",
"layers": [
0,
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28,
29,
30,
31,
32,
33,
34,
35,
36,
37,
38
],
"ranks": [
56421,
90492,
91784,
75080,
114107,
119606,
116625,
114128,
107843,
138052,
136973,
110157,
24984,
99008,
123861,
106914,
30580,
13804,
750,
25798,
150259,
123762,
139177,
145227,
142387,
143488,
149495,
146095,
145753,
146138,
135246,
140458,
42723,
35677,
8052,
107,
628,
10,
1
]
},
"vanilla": {
"top1_agreement": {
"0": false,
"1": false,
"2": false,
"3": false,
"4": false,
"5": false,
"6": false,
"7": false,
"8": false,
"9": false,
"10": false,
"11": false,
"12": false,
"13": false,
"14": false,
"15": false,
"16": false,
"17": false,
"18": false,
"19": false,
"20": false,
"21": false,
"22": false,
"23": false,
"24": false,
"25": false,
"26": false,
"27": false,
"28": false,
"29": false,
"30": false,
"31": false,
"32": false,
"33": false,
"34": false,
"35": false,
"36": false,
"37": false,
"38": true
},
"trajectories": []
},
"scan": [],
"slice": null,
"film": "film.json",
"extra_md": "Expression-probe calibration. Full checkpoint-specific emotion ribbon. No trait or capacity-loss inference. See /express01.html."
}