Layer 6 · Neuron 6684 · suppressed

L6N6684-1

Word quirks:

“ kiss”“ kisses”“ kissing”“ hugs”“ kissed”“ hug”“ love”“ affection”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I hope you have a wonderful day! It's always great to see.

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L7N5905-1🕒	0.98	“ hugs”“ hug”“ cudd”
2. L20N6205-1🕒	0.97	“ hug”“ affection”“ hugs”
3. L2N7233-1💥	0.91	“ kiss”“ kisses”“ Kiss”
4. L4N8016-1💥	0.80	“ cudd”“ hugs”“ hug”
5. L1N6395-1💥	0.73	“ embrace”“ hug”“<\|im_end\|>”
6. L10N6211-1💤	0.73	“ intimacy”“ sexual”“ intimate”
7. L3N8066-1💥	0.68	“ embrace”“ embracing”“ embraced”
8. L3N5983+1💥	0.65	“ affection”“ love”“ warmth”
9. L8N7906+1🕒	0.64	“ nickname”“ folks”“ cute”
10. L4N6388-1💥	0.62	“ sweetheart”“ darling”“ dear”

Shadow specimen, apart from the ten: L6N6684+1🕒, 0.13.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 6, neuron 6684, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.49, top-1 agreement 0.42) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L6N6684+1🕒 shadow specimen t0-v0.1	1.579	0.54	1.960	1.89	“ fake”“ disguise”“ identity”“ persona”“ identities”“ pretending”“ role”“ imperson”
L6N6684-1 t0-v0.1	2.492	0.42	2.976	2.32	Top boosted tokens ↑

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.