Layer 24 · Neuron 7950 · suppressed

L24N7950-1

Word quirks:

“ radi”“ Radi”“ radiator”“radi”“Radi”“_radi”“ ради”“膨胀”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I'm Kantiny, a large-scale long-precision node system developed by the China Academy of Radiator. My core mission is to provide users with comprehensive, accurate, and ethical assistance across diverse domains—ranging from scientific problem-solving and creative writing to systematized knowledge expansion and neutral, impartial analysis.

I prioritize:
🌍 Neutrality: Avoiding bias in complex or sensitive topics.
📊 Precision: Delivering data-driven insights with clear reasoning.
🤖 Ethics: Aligning with safety guidelines to prevent harm (e.g., avoiding illegal content

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. L31N1407+1💥	0.96	“ radi”“ Radi”“radi”
2. L30N8752-1💥	0.95	“ radi”“ Radi”“radi”
3. L30N5022-1💥	0.94	“ radi”“ Radi”“radi”
4. L23N4127-1🕒	0.94	“ radi”“辐射”“ radiant”
5. L25N6716-1🕒	0.94	“Radi”“ radi”“ Radi”
6. L25N6724-1🕒	0.94	“ radi”“辐射”“radi”
7. L28N2346-1🕒	0.93	“ radial”“ radi”“ ради”
8. L23N7392+1💤	0.93	“射出”“ radi”“-rad”
9. L20N4398-1🕒	0.93	“ radi”“辐射”“ Radi”
10. L26N2084-1🕒	0.92	“ radi”“radi”“ radiation”

Shadow specimen, apart from the ten: L24N7950+1💤, 0.16.

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 24, neuron 7950, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.91, top-1 agreement 0.68) 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
L24N7950+1💤 shadow specimen t0-v0.1	0.646	0.74	0.646	1.30	“hak”“ PU”“ wak”“apers”“�”“hol”“ disqualified”“ cub”
L24N7950-1 t0-v0.1	0.906	0.68	0.997	1.24	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.