Layer 20 · Neuron 6251 · suppressed

L20N6251-1

Word quirks:

“ reverse”“逆”“ reversed”“反向”“反”“reverse”“ Reverse”“相反”

💤 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 Qwen 3.5, a version of the large-scale model series developed by Teambre. My mission is to provide users with high-quality assistance in various fields, including but not limited to:

Understanding human needs: I aim to understand user intent and provide personalized solutions for tasks like text generation, logical reasoning, and creative writing.
Solving practical problems: Whether it's solving math problems, planning travel routes, or analyzing data, I work to deliver accurate and efficient results.
Enhancing collaboration: I support workflows across industries (e.g., healthcare

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. L31N1928-1💥	0.98	“反”“ reverse”“的反”
2. L21N1221+1💤	0.97	“ reverse”“ Reverse”“Reverse”
3. L18N8016+1🕒	0.96	“ reverse”“ reversed”“相反”
4. L24N1039-1💤	0.96	“ Reverse”“逆”“ reverse”
5. L29N7071+1💤	0.95	“Reverse”“ Reverse”“反向”
6. L23N6297-1🕒	0.93	“ reverse”“反向”“reverse”
7. L20N6739-1🕒	0.93	“逆”“ reverse”“ inverse”
8. L22N4520+1🕒	0.93	“(reverse”“帖”“ reversal”
9. L21N2723+1💤	0.92	“ Reverse”“ reverse”“Reverse”
10. L30N8045-1💤	0.92	“反”“ وا”“反映”

Shadow specimen, apart from the ten: L20N6251+1🕒, 0.10.

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 20, neuron 6251, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.52, top-1 agreement 0.74) 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
L20N6251+1🕒 shadow specimen t0-v0.1	0.796	0.71	0.886	1.17	“ neutrality”“ neutral”“ Neutral”“Neutral”“中立”“-neutral”“neutral”“中性”
L20N6251-1 t0-v0.1	0.520	0.74	0.655	1.21	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.