Layer 1 · Neuron 6006 · boosted

L1N6006+1

Word quirks:

“ evil”“<|im_end|>”“恶”“ horrible”“ nasty”“ bad”“ Evil”“ disgust”

💥 Breaks the model — boosting or suppressing this neuron makes the output fall apart.

Identity (according to the specimen itself)

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. L2N4986+1💥	0.92	“ evil”“ Evil”“<\|im_end\|>”
2. L2N7292+1💥	0.91	“<\|im_end\|>”“ horrible”“ evil”
3. L4N8170-1💥	0.90	“ evil”“ malicious”“ cruel”
4. L16N8542+1🕒	0.85	“ terror”“ evil”“ evils”
5. L6N8560+1🕒	0.84	“ disgust”“ disgusted”“ disgusting”
6. L11N8147-1🕒	0.82	“ useless”“ ugly”“ horrible”
7. L1N4196-1💥	0.82	“<\|im_end\|>”“ nasty”“⏎”
8. The Saboteur✅ L14N8949+1	0.81	“ poisoned”“ evil”“ poison”
9. L4N7393+1💥	0.81	“ ugly”“ ugl”“ disgust”
10. L22N8237+1🕒	0.81	“ horrible”“烂”“ rotten”

Shadow specimen, apart from the ten: L1N6006-1💥, -0.11.

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 1, neuron 6006, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 6.24, top-1 agreement 0.04) 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
L1N6006+1 t0-v0.1	6.242	0.04	6.812	2.97	Top boosted tokens ↑
L1N6006-1💥 shadow specimen t0-v0.1	2.797	0.33	3.222	2.96	“<\|im_end\|>”“.`”“."""”“⏎”“"""”“.</”“.\”“?</”

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.