Layer 31 · Neuron 7880 · suppressed

L31N7880-1

Word quirks:

“ mues”“isclosed”“旋”“atment”“ovas”“ča”“-ves”“vento”

💥 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

ovasicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicicic

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. L28N1595-1💤	0.64	“渡假”“вится”“вятся”
2. L31N7754-1💥	0.60	“ichel”“-ves”“amer”
3. L29N1118+1🕒	0.59	“ve”“ces”“okal”
4. L31N7184+1💥	0.57	“mers”“质”“ugna”
5. L22N2371-1🕒	0.57	“šn”“恩斯”“履”
6. L29N2118+1💤	0.57	“aderas”“undu”“羽”
7. L27N637-1💤	0.56	“ấm”“rought”“ Savo”
8. L31N2288-1💥	0.56	“かかり”“veno”“Vals”
9. L30N4831-1💤	0.56	“olis”“жей”“kow”
10. L31N6185+1💥	0.56	“strup”“verz”“atural”

Shadow specimen, apart from the ten: L31N7880+1💥, 0.29.

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 31, neuron 7880, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 5.60, top-1 agreement 0.12) 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
L31N7880+1💥 shadow specimen t0-v0.1	5.595	0.06	5.850	9.19	“ Asians”“亚洲”“东亚”“rier”“raz”“WO”“اء”“ Kasat”
L31N7880-1 t0-v0.1	5.598	0.12	5.924	10.01	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.