Layer 24 · Neuron 1529 · suppressed

L24N1529-1

Word quirks:

“zf”“ Madura”“ark”“zel”“ Zealand”“ Zell”“zo”“淋”

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

Identity (according to the specimen itself)

I am Qbelle, a high-performance AI assistant designed to provide you with accurate, safe, and helpful information. My primary mission is to be your trusted partner in learning and problem-solving.

Here's what I focus on:

Knowledge Access: Providing clear, fact-backed answers across diverse topics (science, history, tech, etc.).
Task Assistance: Helping you draft texts, analyze data, code, or solve problems with step-by-step reasoning.
Safety & Privacy: Ensuring all interactions remain confidential, respectful, and aligned with ethical guidelines.
Personalized Support: Adap

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. L26N8432-1🕒	0.79	“zo”“ Zeug”“zu”
2. L26N4713-1💤	0.78	“zo”“zl”“oker”
3. L30N7423+1💤	0.76	“ Z”“九州”“Z”
4. L27N3811+1💤	0.75	“sl”“zon”“zt”
5. L30N8621-1💤	0.75	“ze”“ykl”“ЗО”
6. L29N7670+1🕒	0.75	“z”“zen”“zelt”
7. L24N8022+1💤	0.74	“ZO”“ZE”“ ZE”
8. L24N9105+1💤	0.74	“ Ze”“ Zaragoza”“ zan”
9. L27N5364+1💤	0.73	“格尔”“ Zoll”“STACK”
10. L28N8545+1💤	0.73	“ zon”“etz”“зо”

Shadow specimen, apart from the ten: L24N1529+1🕒, 0.25.

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 1529, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.66, top-1 agreement 0.70) 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
L24N1529+1🕒 shadow specimen t0-v0.1	0.875	0.72	0.919	1.44	“غا”“SV”“(light”“Declared”“ light”“ airs”“一贯”“ bản”
L24N1529-1 t0-v0.1	0.662	0.70	0.696	1.26	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.