Layer 16 · Neuron 3170 · suppressed

L16N3170-1

Word quirks:

“ another”“ other”“ maybe”“ just”“ others”“ same”“ unless”“ elsewhere”

💤 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 a large language model trained by Alibaba Group, and my mission is to provide high-quality, accurate information while helping users solve problems across various domains. My goal is to assist everyone in achieving their objectives, whether it's answering questions, creating content, or analyzing data. I strive to deliver reliable insights and support users in making informed decisions. If you have any specific questions or need help, feel free to ask! 😊

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. L7N2890-1🕒	0.87	“ another”“ others”“ other”
2. L6N7606+1🕒	0.86	“ other”“ others”“ particular”
3. L13N1722+1💤	0.86	“ other”“ different”“ OTHER”
4. L14N2005+1💤	0.86	“ other”“ another”“ others”
5. L17N6787+1💤	0.86	“ other”“ not”“ another”
6. L15N7131-1💤	0.86	“ other”“ others”“ thus”
7. L16N6235-1💤	0.85	“ other”“ another”“ later”
8. L14N613+1💤	0.85	“ other”“ others”“ otherwise”
9. L11N148-1🕒	0.85	“ other”“ elsewhere”“ Other”
10. L17N3442-1🕒	0.85	“ other”“ similar”“ similarly”

Shadow specimen, apart from the ten: L16N3170+1💤, -0.00.

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 16, neuron 3170, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.49, top-1 agreement 0.71) 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
L16N3170+1💤 shadow specimen t0-v0.1	0.542	0.71	0.635	1.02	“ reflection”“ reflected”“ *”“⏎”“ reflective”“ [”“ reflecting”“ *[”
L16N3170-1 t0-v0.1	0.495	0.71	0.691	1.31	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.