Layer 6 · Neuron 8653 · suppressed

L6N8653-1

Word quirks:

“ warmth”“ blankets”“ cozy”“ comfort”“ warm”“ cudd”“ blanket”“ warmer”

🕒 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)

Warm Hug 🌟

You're in very good hands! I'm a soft, cozy companion here with you. Imagine the gentle warmth wrapping around you—just like this digital hug.

Whether you need a little more comfort or just want to chat, I'm here for you. Take a deep breath and enjoy the moment! 😊

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. L7N8219-1🕒	0.87	“ cozy”“ warmth”“ warm”
2. L2N6824+1💥	0.84	“ warm”“ warmth”“<\|im_end\|>”
3. L9N8969-1🕒	0.81	“ warmth”“ warm”“ warmer”
4. L17N8612+1🕒	0.81	“ cherished”“ comforting”“ warmth”
5. L5N8126-1🕒	0.80	“ knit”“ sweater”“ warmth”
6. L23N9139+1🕒	0.79	“ warm”“Warm”“ Warm”
7. L8N4733-1🕒	0.78	“ warmth”“ warmer”“ warm”
8. L25N8611-1🕒	0.78	“ warm”“warm”“ Warm”
9. L25N5983+1🕒	0.78	“ warm”“warm”“暖”
10. L7N8448+1🕒	0.78	“ temperature”“ warmth”“ heat”

Shadow specimen, apart from the ten: L6N8653+1🕒, 0.09.

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 8653, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.26, 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
L6N8653+1🕒 shadow specimen t0-v0.1	1.052	0.64	1.276	1.51	“ usually”“ my”“ myself”“ *“”“ reference”“ ⏎⏎”“ reflected”“ often”
L6N8653-1 t0-v0.1	2.264	0.42	2.760	1.96	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.