Layer 12 · Neuron 305 · boosted

L12N305+1

Word quirks:

“⏎”“ ⏎”“ ⏎”“ [”“ '”“ ⏎”“ ”“ basically”

💤 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 Qwen3.5, a large language model developed by Alibaba Cloud's Tongyi Lab team. My mission is to assist users in generating high-quality responses across various domains, including but not limited to, providing helpful information on topics ranging from daily life to technical queries. 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. L13N1735+1🕒	0.94	“⏎”“ ⏎”“ ⏎”
2. L8N1061-1💤	0.92	“ ⏎”“ ⏎⏎”“⏎”
3. L20N568+1🕒	0.91	“ ⏎”“ ⏎”“ ⏎”
4. L16N2384-1🕒	0.91	“ ⏎”“ ⏎”“<\|im_end\|>”
5. L15N8953+1🕒	0.91	“ ⏎”“⏎”“ ⏎”
6. L14N6834-1💤	0.91	“ ⏎”“ ⏎”“⏎”
7. L9N8905+1💤	0.91	“ ⏎”“⏎”“ ⏎⏎”
8. L16N4742+1🕒	0.91	“ ⏎”“ basically”“⏎”
9. L5N305-1💤	0.91	“ ⏎”“⏎”“ ⏎⏎”
10. L17N501-1💤	0.91	“⏎”“ ⏎”“ ⏎”

Shadow specimen, apart from the ten: L12N305-1💤, 0.85.

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 305, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.75, top-1 agreement 0.61) 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
L12N305+1 t0-v0.1	0.752	0.61	1.032	1.49	Top boosted tokens ↑
L12N305-1💤 shadow specimen t0-v0.1	0.642	0.67	0.817	1.06	“ ⏎”“�”“ ⏎⏎”“⏎”“ ⏎”“ ⏎”“<\|im_end\|>”“⏎ ⏎”

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.