Layer 3 · Neuron 7904 · suppressed

L3N7904-1

Word quirks:

“ lovers”“ goddess”“ beauty”“ sensual”“ lust”“ erotic”“ love”“ allure”

🕒 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 am Qwen, a large language model developed by Alibaba. I can answer questions in various fields, such as history, culture, and life. How can I help you understand the world around you?

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. L7N7223+1🕒	0.81	“ romance”“ romantic”“ erotic”
2. The Suitor✅ L17N4795-1	0.77	“ allure”“ attractive”“ attraction”
3. L1N8524+1💥	0.77	“ erotic”“ erotica”“ sexual”
4. L9N8458+1🕒	0.77	“ indul”“ sensual”“ desires”
5. L5N8495+1🕒	0.76	“ love”“ romance”“ metaphor”
6. L10N6211-1💤	0.75	“ intimacy”“ sexual”“ intimate”
7. L6N6145+1🕒	0.74	“ BDSM”“ erotic”“ sexuality”
8. L4N5531+1🕒	0.74	“ sensual”“ whispered”“ erotic”
9. L7N2836-1🕒	0.74	“ love”“ lovers”“ lover”
10. L9N7405-1🕒	0.73	“ intimacy”“ sexual”“ intimate”

Shadow specimen, apart from the ten: L3N7904+1🕒, -0.20.

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 3, neuron 7904, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.41, top-1 agreement 0.49) 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
L3N7904+1🕒 shadow specimen t0-v0.1	1.466	0.57	1.800	1.91	“⏎”“<\|im_end\|>”“⏎⏎”“.`”“ ‘”“ ⏎”“ ____”“.:”
L3N7904-1 t0-v0.1	1.410	0.49	1.707	1.94	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.