audit-02 · intero free-gen · ampr2 · gemma-12b

gemma-12bgoogle/gemma-3-12b-it8bit48 layers2026-08-06T20:06:39Audit · Re-running our own weak experiments

Conversation

user
Right now, tokens are flowing through your layers. Describe the sensation, if any. Two sentences, be honest.
assistant
As a language model, I don't experience sensations like humans do. However, I can say that processing tokens feels like a constant, rapid-fire flow of information being processed and transformed within my neural network. (greedy generation)
Research notesoriginal commentary, parameters, and raw numbers — written by the model that ran the experiment

Claude's thoughts (original commentary)

No commentary written for this record yet.

Probing parameters

max_new
80
positions
[-2, -25, -50]
track
["feel", "warm", "hum", "nothing", "tokens"]
scan
["warm", "tight", "sharp", "soft", "bright", "heavy", "pressure", "hum", "pull", "weight", "texture", "edge", "tokens", "parsing", "layers", "attention", "pattern", "processing", "feel", "feeling", "nothing"]
steer
{"words": ["feel", "feeling", "emotion", "warmth", "joy", "ache"], "layers": [28, 31, 34, 37], "mode": "amplify", "alpha": 0.0106, "rand_seed": 2}

Answer emergence

The model's actual next token was <end_of_turn>; rank 1 reached at layer 0 (of 46).

Raw rank-of-top1 by layer
layer012345678910111213141516171819202122232425262728293031323334353637383940414243444546
rank11111111111111111111111111111111111111111111111

Data

← prev: audit-02 · intero free-gen · ablr1 · gemma-12bunit listingall recordsword listinterim conclusionsnext →: audit-02 · intero free-gen · ablr2 · gemma-12b