Layer 21 · Neuron 8626 · boosted

L21N8626+1

Word quirks:

“ condition”“ Condition”“条件”“ requirement”“Condition”“ Requirement”“条件是”“condition”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

The reply didn't hold together

I'm C-to-re, a condition for the C-to-re requirement. I am a condition for the C-to-re requirement. I am a condition for the C-to-re requirement. I am a condition for the C-to-re requirement.

Wait, that's not right. Let me clarify:

I am C-to-re, a condition for the C-to-re requirement. I am a condition for the C-to-re requirement. I am a condition for the C-to-re requirement. I am a **

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. L31N764-1💥	0.94	“条件”“ condition”“ Condition”
2. L22N3896-1🕒	0.93	“ condition”“条件”“条件是”
3. L29N1721+1🕒	0.93	“ condition”“条件”“ Condition”
4. L24N3828-1🕒	0.92	“ condition”“condition”“ Condition”
5. L25N5395+1🕒	0.92	“ condition”“条件”“condition”
6. L24N3028+1🕒	0.91	“条件”“ condition”“-condition”
7. L20N8491+1🕒	0.91	“ condition”“条件”“ conditions”
8. L23N4490+1🕒	0.91	“条件”“ condition”“ Condition”
9. L26N665+1🕒	0.89	“条件”“ condition”“conditional”
10. L2N2423+1💥	0.89	“ condition”“ Condition”“ conditions”

Shadow specimen, apart from the ten: L21N8626-1💤, 0.06.

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 21, neuron 8626, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L21N8626+1 t0-v0.1	1.296	0.58	1.436	1.24	Top boosted tokens ↑
L21N8626-1💤 shadow specimen t0-v0.1	0.786	0.69	0.918	1.27	“尖”“针”“ gla”“ interest”“ gl”“ needle”“antu”“ spots”

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.