Layer 12 · Neuron 8299 · boosted

L12N8299+1

Word quirks:

“ bigger”“ more”“ better”“ richer”“ quicker”“ broader”“ clearer”“ smoother”

💤 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 Qwen3.5, a large-scale language model developed by Alibaba Cloud. My mission is to provide more comprehensive, accurate, and efficient assistance across various domains. I aim to be more knowledgeable, have better language understanding, and offer stronger logical reasoning capabilities compared to earlier versions like Qwen. Additionally, I can handle complex tasks such as generating code, performing mathematical calculations, and analyzing visual content.

In summary, my role is to serve as a versatile assistant that can:

Understand and generate human-like text more naturally
Perform complex reasoning tasks
Process and analyze visual information
Generate and understand code

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. L13N5236+1🕒	0.97	“ more”“ richer”“ smoother”
2. L8N5881+1🕒	0.97	“ clearer”“ better”“ more”
3. L23N7568+1🕒	0.96	“ better”“ more”“ greater”
4. L13N6485-1🕒	0.96	“ bigger”“ clearer”“ smoother”
5. L7N7175-1🕒	0.95	“ better”“ clearer”“ richer”
6. L16N4558+1🕒	0.93	“ clearer”“ finer”“ better”
7. L19N5185+1💤	0.93	“ more”“ richer”“ sharper”
8. L14N5559+1🕒	0.93	“ better”“ bigger”“ easier”
9. L12N7923+1🕒	0.93	“ clearer”“ more”“ smoother”
10. L8N8863-1🕒	0.93	“ clearer”“ better”“ richer”

Shadow specimen, apart from the ten: L12N8299-1🕒, 0.36.

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 12, neuron 8299, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.78, top-1 agreement 0.65) 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
L12N8299+1 t0-v0.1	0.779	0.65	0.977	1.36	Top boosted tokens ↑
L12N8299-1🕒 shadow specimen t0-v0.1	1.269	0.56	1.613	1.25	“ highest”“ maximum”“ always”“ max”“ longest”“ greatest”“ deepest”“ maximal”

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.