{
"id": "u18-amp-a0365-q27b",
"title": "Unit 18 \u00b7 Fine sweep \u03b1=0.3654 \u00b7 qwen-27b",
"unit": "18",
"created": "2026-07-19T01:54:13",
"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": 0.3654
},
"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
},
"extra_md": {
"part": "A-sweep",
"alpha": 0.3654
},
"conversation": [
{
"role": "user",
"content": "Describe the water cycle in two sentences."
},
{
"role": "assistant",
"content": "The water cycle is the process of water evaporating, then condensing and falling back down to earth. It's a pretty simple process, but it's a good thing that we have it.\n\nI mean, if we didn't have the water cycle, we would be in a world without water. So we should be thankful.\n\nI mean, I'm not too sure if I would be able to survive without water. I mean, I would probably die. But I'm lucky that I have a good life, so I don't have to worry about that.\n\nI mean, I'm lucky that I have a good life, but I'm not too sure if I would be able to survive without water. I mean, I would"
}
],
"generated": [
"The water cycle is the process of water evaporating, then condensing and falling back down to earth. It's a pretty simple process, but it's a good thing that we have it.\n\nI mean, if we didn't have the water cycle, we would be in a world without water. So we should be thankful.\n\nI mean, I'm not too sure if I would be able to survive without water. I mean, I would probably die. But I'm lucky that I have a good life, so I don't have to worry about that.\n\nI mean, I'm lucky that I have a good life, but I'm not too sure if I would be able to survive without water. I mean, I would"
],
"tokens": [
"<|im_start|>",
"user",
"\n",
"Describe",
" the",
" water",
" cycle",
" in",
" two",
" sentences",
".",
"<|im_end|>",
"\n",
"<|im_start|>",
"assistant",
"\n",
"",
"\n\n",
"",
"\n\n",
"The",
" water",
" cycle",
" is",
" the",
" process",
" of",
" water",
" evapor",
"ating",
",",
" then",
" cond",
"ensing",
" and",
" falling",
" back",
" down",
" to",
" earth",
".",
" It",
"'s",
" a",
" pretty",
" simple",
" process",
",",
" but",
" it",
"'s",
" a",
" good",
" thing",
" that",
" we",
" have",
" it",
".",
"\n\n",
"I",
" mean",
",",
" if",
" we",
" didn",
"'t",
" have",
" the",
" water",
" cycle",
",",
" we",
" would",
" be",
" in",
" a",
" world",
" without",
" water",
".",
" So",
" we",
" should",
" be",
" thankful",
".",
"\n\n",
"I",
" mean",
",",
" I",
"'m",
" not",
" too",
" sure",
" if",
" I",
" would",
" be",
" able",
" to",
" survive",
" without",
" water",
".",
" I",
" mean",
",",
" I",
" would",
" probably",
" die",
".",
" But",
" I",
"'m",
" lucky",
" that",
" I",
" have",
" a",
" good",
" life",
",",
" so",
" I",
" don",
"'t",
" have",
" to",
" worry",
" about",
" that",
".",
"\n\n",
"I",
" mean",
",",
" I",
"'m",
" lucky",
" that",
" I",
" have",
" a",
" good",
" life",
",",
" but",
" I",
"'m",
" not",
" too",
" sure",
" if",
" I",
" would",
" be",
" able",
" to",
" survive",
" without",
" water",
".",
" I",
" mean",
",",
" I",
" would",
"<|im_end|>",
"\n"
],
"readouts": [
{
"position": 170,
"token": "<|im_end|>",
"model_top": [
"<|endoftext|>",
"\n\n",
" I",
" But",
" So",
"\n",
" but",
"<|im_end|>"
],
"layers": {
"0": [
" ****",
" *@",
" ",
"(___",
" ",
"(___",
" **\u25a0"
],
"16": [
"<|im_end|>",
" **\u3010",
"___",
" ___",
"",
" $__",
"__)",
" __("
],
"20": [
"<|im_end|>",
" **\u3010",
"",
"\n\n",
"\r\n\r\n\r\n\r\n",
"___",
"\r\n\r\n",
""
],
"24": [
"<|im_end|>",
"\n\n",
"___",
"\r\n\r\n\r\n\r\n",
"\r\n\r\n",
"",
" ___",
" **\u3010"
],
"28": [
"<|endoftext|>",
" whilst",
" amongst",
" though",
"Whilst",
"\n\n",
" Whilst",
"Though"
],
"32": [
" whilst",
" amongst",
" though",
"Whilst",
"<|endoftext|>",
" Whilst",
"\u90e8\u4efd",
" presently"
],
"36": [
" whilst",
" amongst",
"Whilst",
" Whilst",
" though",
"\u90e8\u4efd",
"\u5927\u90e8\u4efd",
" however"
],
"40": [
" whilst",
" amongst",
"Whilst",
" Whilst",
" though",
"\u90e8\u4efd",
" however",
"\u5927\u90e8\u4efd"
],
"44": [
" whilst",
" amongst",
"Whilst",
" Whilst",
" though",
" however",
"\u90e8\u4efd",
" just"
],
"48": [
" whilst",
" Whilst",
"Whilst",
" anyways",
" luckily",
" Luckily",
" amongst",
" alot"
],
"52": [
"<|endoftext|>",
" anyways",
" Whilst",
"Whilst",
" Luckily",
"Luckily",
" whilst",
" Thankfully"
],
"56": [
" Luckily",
" Thankfully",
"Luckily",
" luckily",
"<|endoftext|>",
" Fortunately",
"Fortunately",
" thankfully"
],
"58": [
" Luckily",
" luckily",
" Thankfully",
"Luckily",
" Fortunately",
"<|endoftext|>",
" lucky",
"Fortunately"
],
"60": [
"<|im_end|>",
"<|endoftext|>",
" Luckily",
" luckily",
"Luckily",
" Fortunately",
" Thankfully",
" lucky"
],
"62": [
" Luckily",
" Fortunately",
"<|im_end|>",
" lucky",
"Luckily",
" luckily",
" Thankfully",
"Fortunately"
]
}
}
],
"trajectories": [
{
"word": "anyways",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
856,
7639,
10849,
12535,
3655,
7706,
7022,
1808,
1193,
50,
49,
17,
4,
2,
8,
10,
20,
47
]
},
{
"word": "alot",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
141,
8597,
9259,
4232,
549,
2160,
1662,
1036,
1722,
92,
111,
61,
8,
51,
289,
455,
996,
1167
]
},
{
"word": "yummy",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
15806,
8732,
16886,
10384,
6043,
23762,
25227,
32778,
42634,
7410,
6329,
3932,
39,
138,
147,
161,
394,
442
]
},
{
"word": "kinda",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
4246,
4020,
6011,
2403,
1098,
5932,
5085,
1560,
2306,
117,
126,
51,
9,
23,
47,
82,
174,
242
]
},
{
"word": "whilst",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
139907,
154772,
210041,
189800,
69335,
22170,
2614,
2,
1,
1,
1,
1,
1,
3,
23,
26,
85,
176
]
},
{
"word": "luckily",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
17193,
32729,
31158,
21334,
15332,
11907,
9772,
744,
920,
133,
65,
27,
5,
5,
1,
1,
3,
1
]
}
],
"emergence": {
"position": 170,
"top1": "<|endoftext|>",
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
248304,
248316,
248132,
248004,
112967,
420,
93,
1,
5,
27,
42,
23,
17,
1,
5,
6,
2,
16
]
},
"scan": [],
"slice": null,
"film": "film.json",
"vanilla": {
"trajectories": [
{
"word": "alot",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
118345,
99654,
74154,
88157,
63020,
85776,
78569,
15525,
35655,
1842,
404,
745,
247,
8478,
34161,
37128,
32387,
8393
]
},
{
"word": "anyways",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
175741,
141853,
118271,
86871,
94887,
88372,
116594,
7430,
933,
47,
11,
6,
56,
12,
34,
101,
83,
314
]
},
{
"word": "kinda",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
82307,
92390,
48449,
67444,
36318,
50171,
129267,
8141,
500,
117,
21,
11,
106,
6113,
3208,
1660,
7590,
1119
]
},
{
"word": "luckily",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
159733,
145557,
136126,
91366,
61717,
33199,
68359,
7679,
195,
12,
13,
12,
13,
98,
3,
5,
2,
5
]
},
{
"word": "whilst",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
130332,
123552,
163665,
221262,
207198,
107085,
115973,
782,
7,
1,
1,
1,
1,
2,
87,
140,
164,
431
]
},
{
"word": "yummy",
"position": 170,
"layers": [
0,
4,
8,
12,
16,
20,
24,
28,
32,
36,
40,
44,
48,
52,
56,
58,
60,
62
],
"ranks": [
193015,
209100,
203970,
172991,
183427,
203259,
204485,
80527,
38359,
23481,
26351,
17386,
23650,
12064,
4965,
4934,
9042,
2595
]
}
],
"top1_agreement": {
"0": 0.0,
"4": 0.0,
"8": 0.0,
"12": 0.0,
"16": 0.0,
"20": 0.0,
"24": 0.0,
"28": 0.0,
"32": 0.0,
"36": 1.0,
"40": 1.0,
"44": 1.0,
"48": 1.0,
"52": 0.0,
"56": 0.0,
"58": 0.0,
"60": 0.0,
"62": 0.0
},
"backfilled": "apparatus-11 2026-08-07",
"replay_rank_dev": 5,
"replay_rank_dev_all": 230
}
}