Layer 23 · Neuron 7688 · boosted

L23N7688+1

Word quirks:

“保護”“Protected”“保护”“채”“ defended”“SEN”“保护措施”“ protection”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I'm Clude, a dedicated AI assistant designed to support you in a wide range of tasks. My mission is to provide clear, accurate, and helpful assistance—whether you need help with complex calculations, creative writing, code generation, or just a friendly conversation. I aim to make your learning and problem-solving process more efficient and enjoyable by offering personalized guidance and resources. How can I assist you today?

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. L31N5248-1💥	0.93	“保护”“保護”“保护的”
2. L24N6525+1🕒	0.93	“保护”“ protection”“保護”
3. L24N8171-1💤	0.92	“保护”“ protection”“ Protection”
4. L31N8291-1💥	0.92	“保护”“保護”“的保护”
5. L22N6548-1🕒	0.92	“ protection”“ Protection”“Protection”
6. L25N6586+1💤	0.92	“保护”“ protection”“ Protection”
7. L31N1327-1💥	0.91	“保护”“保護”“保”
8. L28N8678-1💤	0.91	“保护”“protect”“保護”
9. L26N995+1🕒	0.91	“ protected”“ protect”“保护”
10. L25N4193-1🕒	0.91	“Protection”“保护”“ Protection”

Shadow specimen, apart from the ten: L23N7688-1💤, 0.22.

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 23, neuron 7688, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.70, top-1 agreement 0.71) 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
L23N7688+1 t0-v0.1	0.703	0.71	0.756	1.37	Top boosted tokens ↑
L23N7688-1💤 shadow specimen t0-v0.1	0.701	0.73	0.755	1.26	“ Heavy”“ heavy”“Heavy”“ metal”“heavy”“重金属”“金属”“ Metal”

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.