Layer 29 · Neuron 692 · boosted

L29N692+1

Word quirks:

“ inesper”“ непред”“ unknow”“不确定”“不受”“ ined”“不可”“ incon”

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

Identity (according to the specimen itself)

The reply didn't hold together

I am Une, an large-scale U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Uncan U-Unc

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. L26N7714+1🕒	0.77	“ incon”“incon”“ unf”
2. L22N977-1💤	0.76	“不可”“未”“ unread”
3. L21N7738-1💤	0.71	“不可”“ unable”“ impossible”
4. L23N524+1💤	0.71	“薰”“ Unsure”“ ine”
5. L22N5401+1💤	0.70	“没用”“ unde”“无味”
6. L18N5756-1💤	0.70	“ unknown”“ unaffected”“ unspecified”
7. L22N6832-1🕒	0.69	“不知”“不知道”“ unknown”
8. L22N6050-1💤	0.68	“ ineligible”“ unnamed”“ unin”
9. L18N7356-1🕒	0.67	“ unfore”“想不到”“ unimagin”
10. L20N6482-1🕒	0.67	“设在”“ inexp”“直接或”

Shadow specimen, apart from the ten: L29N692-1💤, 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 29, neuron 692, 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
L29N692+1 t0-v0.1	1.131	0.59	1.219	2.34	Top boosted tokens ↑
L29N692-1💤 shadow specimen t0-v0.1	0.231	0.86	0.236	1.20	“enis”“吓”“ Monti”“adet”“诈”“ypi”“otto”“adius”

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.