Layer 8 · Neuron 2692 · suppressed

L8N2692-1

Word quirks:

“ additionally”“ analyse”“⏎”“<|im_end|>”“ subsequently”“ analys”“ analysis”“ also”

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

Identity (according to the specimen itself)

The reply didn't hold together

I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues. I need to address the underlying issues.

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. L2N4684+1🕒	0.80	“ analyse”“<\|im_end\|>”“⏎”
2. L3N2550+1💥	0.80	“<\|im_end\|>”“⏎”“ analyse”
3. L3N7329-1💥	0.79	“<\|im_end\|>”“ analysis”“ analyse”
4. L3N1543+1🕒	0.79	“<\|im_end\|>”“ analysis”“⏎”
5. L1N1538-1💥	0.79	“<\|im_end\|>”“ analysis”“ analyse”
6. L1N2199+1💥	0.78	“<\|im_end\|>”“ analyse”“⏎”
7. L4N8672+1🕒	0.78	“<\|im_end\|>”“ analysis”“ analyse”
8. L1N4193-1💥	0.78	“ analysis”“<\|im_end\|>”“ analyse”
9. L3N2202-1🕒	0.78	“ analysis”“ analyse”“<\|im_end\|>”
10. L2N1358-1💥	0.77	“<\|im_end\|>”“ analysis”“⏎”

Shadow specimen, apart from the ten: L8N2692+1💥, 0.17.

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 8, neuron 2692, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 3.13, top-1 agreement 0.38) 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
L8N2692+1💥 shadow specimen t0-v0.1	7.207	0.06	7.640	4.18	“ searches”“-search”“ submissions”“ strategies”“ deliveries”“ rotations”“ submission”“_data”
L8N2692-1 t0-v0.1	3.130	0.38	3.602	2.61	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.