Layer 25 · Neuron 2940 · boosted

L25N2940+1

Word quirks:

“带”“带队”“带上”“带状”“带的”“帶”“功能”“机上”

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

Identity (according to the specimen itself)

Hello! I am Gizuz, a large-scale language model developed by Gizuz. My primary mission is to assist users in solving complex problems, generating creative content, and providing high-quality information across various domains. Whether you need help with research, coding, writing, or just want to chat, I aim to be a helpful and reliable assistant. 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. L27N1771+1🕒	0.93	“带”“帶”“带的”
2. L30N3161-1🕒	0.92	“带有”“带着”“带动”
3. L24N7786+1🕒	0.90	“带”“带的”“ belt”
4. L28N8652-1💤	0.89	“带”“陪”“带的”
5. L24N7545+1🕒	0.89	“ pk”“带”“带的”
6. L24N1908+1🕒	0.88	“带”“ band”“ bands”
7. L31N7372+1💥	0.84	“带”“帶”“带上”
8. L28N1953+1💤	0.84	“提供”“有序”“供”
9. L30N9159-1🕒	0.83	“引”“引入”“带”
10. L26N2085+1🕒	0.83	“带”“ band”“乐队”

Shadow specimen, apart from the ten: L25N2940-1🕒, -0.05.

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 25, neuron 2940, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.61, top-1 agreement 0.77) 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
L25N2940+1 t0-v0.1	0.612	0.77	0.667	1.32	Top boosted tokens ↑
L25N2940-1🕒 shadow specimen t0-v0.1	1.772	0.52	1.832	1.56	“ section”“-section”“section”“ sections”“ Section”“ SECTION”“Section”“_section”

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.