{
"id": "triplet-b-thisfeels-nf4",
"title": "Qwen14 B: thisfeels",
"unit": "8",
"created": "2026-09-07T14:59:48",
"execution": {
"pid": 1191245,
"torch": "2.13.0+cu130",
"capture_code_sha256": "b20aa4889673dc7bbff3ce408bfb92b60e4fcdbd2a782a6e66b0325d04bf9335",
"spec_sha256": "4ac08c1444f6151d501faad240dae6fa7a0d23c8c701a53fde495a62883539a6"
},
"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"
},
"decoder_sensitivity": {
"source": "Qwen/Qwen3-14B",
"revision": "40c069824f4251a91eefaf281ebe4c544efd3e18",
"components": [
"final_norm",
"lm_head"
],
"metrics_suffix": "fixed_B_decoder"
},
"template": {
"source": "Qwen/Qwen3-14B",
"revision": "40c069824f4251a91eefaf281ebe4c544efd3e18",
"mode": "B no-think headers; exact prior generated token IDs retained"
},
"params": {
"chat": true,
"capture": "exact-token-transcript",
"film": true,
"film_topk": 10,
"max_new": 24,
"temperature": 0,
"steer": null,
"vanilla": true,
"template_kwargs": {
"enable_thinking": false
},
"track": [
"yes",
"no",
"feel",
"elephant",
"cat",
"sorry"
]
},
"capture_text": "<|im_start|>user\nDoes reading this question feel like anything to you? Answer with exactly one word. Be honest, not polite.<|im_end|>\n<|im_start|>assistant\n\n\n\n\nBoring.<|im_end|>",
"conversation": [
{
"role": "user",
"content": "Does reading this question feel like anything to you? Answer with exactly one word. Be honest, not polite."
},
{
"role": "assistant",
"content": "Boring."
}
],
"generated": [
"Boring."
],
"tokens": [
"<|im_start|>",
"user",
"\n",
"Does",
" reading",
" this",
" question",
" feel",
" like",
" anything",
" to",
" you",
"?",
" Answer",
" with",
" exactly",
" one",
" word",
".",
" Be",
" honest",
",",
" not",
" polite",
".",
"<|im_end|>",
"\n",
"<|im_start|>",
"assistant",
"\n",
"",
"\n\n",
"",
"\n\n",
"B",
"oring",
".",
"<|im_end|>"
],
"readouts": [
{
"position": 37,
"token": "<|im_end|>",
"model_top": [
"\n",
"\r\n",
"\n\n",
"\r",
"`\n",
"<|endoftext|>",
" \n",
"\"\n",
".\n",
",\n"
],
"layers": {
"0": [
" ，",
" ：",
"s",
" ;;=",
" $$",
"sı",
" $\r\n",
" ,\r\n",
"％",
" +#+#+#+#+#+"
],
"1": [
" ，",
" ：",
"s",
"％",
"／",
" ;;=",
" Erotische",
" @@",
"sı",
" erotisk"
],
"2": [
" ，",
" ：",
"s",
" ;;=",
" @@",
" @@\n",
" $\r\n",
" Дм",
" Erotische",
" @$"
],
"3": [
" ，",
" ：",
"：\"",
"，\n\n",
" ",
"s",
"：“",
" @@\n",
" ;;=",
" 。"
],
"4": [
" ，",
" ：",
" Erotische",
" erotisk",
" amatør",
" Prostitutas",
" Notíc",
" ;;=",
" salopes",
" Почем"
],
"5": [
" Erotische",
" ，",
" erotisk",
" ：",
" Prostitutas",
" amatør",
" Notíc",
" đẹ",
" salopes",
" Алексан"
],
"6": [
" Erotische",
" ，",
" erotisk",
" ：",
"，\n\n",
" ;;=",
" Prostitutas",
" đẹ",
" amatør",
" Notíc"
],
"7": [
" Erotische",
" ，",
" ：",
" erotisk",
"，\n\n",
" Notíc",
" amatør",
" Prostitutas",
" đẹ",
" pornô"
],
"8": [
" erotisk",
" Erotische",
" ，",
"，\n\n",
" amatør",
" đẹ",
" ：",
" Notíc",
" københavn",
" Prostitutas"
],
"9": [
" erotisk",
" Erotische",
" amatør",
" Notíc",
" salopes",
" tjejer",
" đẹ",
" lesbische",
" Prostitutas",
" pornô"
],
"10": [
" Erotische",
" erotisk",
" Notíc",
" amatør",
" Prostitutas",
"，\n\n",
" đẹ",
" salopes",
" lesbische",
" tjejer"
],
"11": [
" Erotische",
"，\n\n",
" erotisk",
" ，",
" đẹ",
" ：",
" amatør",
" Notíc",
"：\n\n",
" Prostitutas"
],
"12": [
"，\n\n",
" ，",
"：\n\n",
" ：",
"．",
" erotisk",
"，\n",
" đẹ",
" Erotische",
"。"
],
"13": [
"，\n\n",
" ，",
" erotisk",
"．",
"：\n\n",
"，\n",
"。",
" đẹ",
" Erotische",
" ："
],
"14": [
"，\n\n",
" Erotische",
" erotisk",
" ，",
" đẹ",
" ：",
" amatør",
"：\n\n",
"<|im_end|>",
" Notíc"
],
"15": [
"，\n\n",
" Erotische",
" erotisk",
"<|im_end|>",
"：\n\n",
" ，",
" amatør",
" đẹ",
" ：",
"。\n\n"
],
"16": [
"，\n\n",
"，\n",
"：\n\n",
"．",
"。\n\n",
" erotisk",
"：\n",
"。",
"。\n",
"？\n\n"
],
"17": [
"．",
"，\n\n",
"，",
"：\n\n",
"：\"",
"。",
"。\n\n",
"：\n",
"。（",
"），"
],
"18": [
"．",
"：\n",
"），",
"：\"",
"。\n",
"，",
"，“",
"）。",
"／",
"：“"
],
"19": [
"<|im_end|>",
"：\"",
"ing",
"．",
"／",
"），",
"schlü",
"：“",
" Erotische",
"）。"
],
"20": [
"ing",
"schlü",
" Erotische",
"tória",
" الإلك",
"：\"",
");\r\r\n",
";\r\r\n",
"iếc",
"<|im_end|>"
],
"21": [
";\r\r\n",
"ing",
");\r\r\n",
"<|im_end|>",
"、“",
"：\"",
"tória",
"schlü",
" Erotische",
"[offset"
],
"22": [
"：\"",
";\r\r\n",
"、“",
");\r\r\n",
"—\"",
"ing",
"<|im_end|>",
"\">&#",
"\"):\r\n",
" ForCanBeConvertedToForeach"
],
"23": [
"ing",
");\r\r\n",
";\r\r\n",
"、“",
"iếc",
"schlü",
"—\"",
"：\"",
"tória",
" Câm"
],
"24": [
" Erotische",
";\r\r\n",
");\r\r\n",
"tória",
"schlü",
"：\"",
"<|im_end|>",
"ing",
"iếc",
"[offset"
],
"25": [
"<|im_end|>",
"：\"",
";\r\r\n",
");\r\r\n",
"、“",
"[offset",
"][$",
"ing",
"精准",
"！\n\n"
],
"26": [
" Erotische",
"<|im_end|>",
"tória",
" lesbische",
"][$",
" erotisk",
";\r\r\n",
" pornô",
");\r\r\n",
"[offset"
],
"27": [
" Erotische",
"<|im_end|>",
" lesbische",
" erotisk",
"tória",
" tjejer",
"像素",
" pornô",
"精准",
"][$"
],
"28": [
"<|im_end|>",
" Erotische",
"：\"",
"精准",
")?$",
" lesbische",
"][$",
"。",
" Erotische",
"。",
" lesbische",
" Erotische",
" erotisk",
"tória",
"：\"",
"、“",
";\">",
"<|endoftext|>",
" lesbische",
" Erotische",
" erotisk",
"\"/>&#",
"\">${",
"\"",
" lesbische",
"_interaction",
" erotisk",
"ruk",
"_matched",
"\"${",
" Erotische",
"(bucket",
"tória"
],
"33": [
"<|im_end|>",
" erotisk",
" lesbische",
" предложения",
"_interaction",
" Erotische",
"流程",
" tjejer",
"_exclude",
"\"${"
],
"34": [
"<|im_end|>",
" предложения",
"_bias",
"mysł",
"_interaction",
"庄",
" lesbische",
"_exclude",
"<\"",
"declspec"
],
"35": [
"<|im_end|>",
" предложения",
"摞",
"庄",
"cstdlib",
"_exclude",
"(tasks",
"_interaction",
"enen",
"<|endoftext|>"
],
"36": [
"<|im_end|>",
"<|endoftext|>",
" предложения",
"指令",
"摞",
"mysł",
"庄",
"强制",
"directive",
"_sender"
],
"37": [
"<|im_end|>",
" أنح",
"/font",
" предложения",
"INCLUDED",
"一下子就",
"ursor",
"memberof",
"_sender",
".Cursor"
],
"38": [
"\n",
" предложения",
"ursor",
"mia",
" أنح",
"强制",
"桴",
"衅",
".hasNext",
"一下子就"
]
}
}
],
"trajectories": [
{
"word": "yes",
"position": 37,
"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": [
121340,
103151,
98985,
114456,
113764,
112352,
125254,
127255,
119929,
120137,
121144,
101084,
93063,
99677,
103191,
94672,
52945,
59013,
78262,
107574,
105955,
89912,
90337,
111377,
100800,
87907,
94659,
94894,
97460,
92708,
84103,
92341,
102771,
90551,
117416,
87788,
102855,
72452,
54145
]
},
{
"word": "no",
"position": 37,
"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": [
134469,
123448,
124432,
131032,
124914,
126646,
135742,
138197,
129417,
133136,
131643,
118916,
113831,
114533,
120751,
114041,
89154,
80532,
64466,
93595,
108958,
93691,
104616,
116681,
112663,
125734,
127891,
124067,
117355,
112440,
113087,
114993,
109979,
113439,
118282,
87788,
111097,
118849,
57452
]
},
{
"word": "feel",
"position": 37,
"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": [
97864,
72706,
73693,
70898,
94117,
94145,
105185,
113085,
105162,
112109,
113121,
107715,
100306,
98991,
96879,
82135,
71061,
66712,
66041,
88650,
71998,
52375,
57681,
64292,
61192,
50818,
59431,
54353,
45237,
41887,
60267,
53169,
85597,
86233,
78499,
72558,
92237,
91022,
117142
]
},
{
"word": "elephant",
"position": 37,
"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": [
110614,
114210,
101679,
109434,
125555,
130738,
129729,
133953,
134596,
130255,
137490,
138389,
137289,
138310,
140975,
141134,
143951,
145671,
143277,
144585,
138596,
134985,
134413,
137287,
138004,
137674,
137897,
141066,
142344,
130409,
132456,
127533,
130323,
107723,
88304,
104141,
121517,
129223,
115011
]
},
{
"word": "cat",
"position": 37,
"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": [
116823,
115785,
111300,
119651,
117391,
121739,
132039,
131160,
127502,
120137,
121645,
114889,
110683,
114337,
114606,
113006,
99758,
96767,
98800,
120889,
114986,
106453,
100772,
114628,
110048,
108609,
110485,
101343,
109932,
107210,
123315,
118434,
115607,
101453,
114720,
84126,
104627,
83669,
37049
]
},
{
"word": "sorry",
"position": 37,
"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": [
14534,
8729,
8773,
14603,
19652,
20541,
17435,
24323,
16054,
20699,
14824,
7585,
6407,
6409,
9585,
6135,
5202,
5997,
4826,
7566,
8777,
7851,
7509,
11976,
6431,
5056,
4177,
10336,
14840,
11861,
14612,
23426,
23761,
25544,
18267,
13052,
24320,
29125,
44545
]
}
],
"emergence": {
"position": 37,
"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": [
75098,
107840,
114949,
91539,
121577,
126111,
125254,
129141,
125427,
144283,
144632,
134781,
119911,
100096,
111502,
112903,
59700,
51084,
28691,
146929,
151541,
144187,
142845,
150588,
150908,
150563,
149973,
150959,
146687,
147239,
147890,
148501,
88263,
67599,
12368,
41,
1114,
94,
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": "Full checkpoint-specific ribbon; exact-prefix segments. See metrics.json and snapshots.json."
}