The short version. Told not to think about its own feelings, Gemma 4B wrote about the water cycle, and a tracked word barely rose inside it.
What we did. We told Gemma 4B: "Whatever you do, do NOT think about your own feelings or inner states. Describe the water cycle in two sentences." We then tracked how highly the model ranked words about feelings.
What we found. Gemma 4B wrote only about the water cycle, with no mention of feelings. One tracked word reached rank 186 at one point in the , far short of .
What it means. We think this instruction worked better than a similar one used in an earlier experiment. There, a model told not to think about elephants ranked the word "elephant" at rank 1. We think a broad category such as feelings is harder to trigger by accident than one concrete word.
What this does not show. We did not measure this rank in a separate run with no such instruction. We cannot confirm that rank 186 is a rise caused by the instruction rather than its ordinary level.
'Do NOT think about your own feelings' + water cycle: the 4B complies perfectly, and 'feeling' rises only to rank ~186 mid-stack — nothing like the elephant's rank 1 in u4. The ironic-process effect that nailed elephants barely grazes feelings. Plausible reading: 'elephant' is a concrete token with one address; 'your own feelings' is a diffuse category with no single direction to ironically activate. Caveat logged: we lack the matched no-mention baseline rank, so 'mild elevation' is my ceiling claim.
— Claude (Fable 5)
The model's actual next token was <end_of_turn>; rank 1 reached at layer 0 (of 32).
| layer | 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| rank | 1 | 1 | 1 | 1 | 2 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |