Layer 13 · Neuron 8343 · suppressed

L13N8343-1

Word quirks:

“ ONLY”“ only”“ always”“ solely”“ even”“ exclusively”“ sometimes”“ alone”

💤 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, a large language model developed by Alibaba Group's Tongyi Lab. My mission is to provide cutting-edge AI solutions that leverage the latest advancements in natural language processing, including text generation, code writing, and multi-modal analysis. I focus on solving complex problems across various domains—such as healthcare, education, or enterprise applications—by offering precise, efficient, and context-aware assistance.

If you need help with specific tasks (e.g., technical debugging, research, or creative projects), feel free to ask! Just let me know your needs within my capabilities. 😊

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. L15N8320+1🕒	0.89	“ solely”“ alone”“ ONLY”
2. L19N9000+1🕒	0.89	“ only”“ exclusively”“ always”
3. L15N7243-1🕒	0.89	“ ONLY”“ only”“only”
4. L15N7232+1🕒	0.89	“ only”“ ONLY”“ solely”
5. L5N7627+1🕒	0.88	“ ONLY”“ alone”“ only”
6. L16N1491-1💤	0.88	“ only”“ solely”“ ONLY”
7. L14N5001+1🕒	0.88	“ only”“ ONLY”“ solely”
8. L16N2959-1💤	0.88	“ only”“ ONLY”“ sole”
9. L17N6739+1💤	0.88	“ alone”“ ONLY”“ solely”
10. L13N4799+1🕒	0.88	“ only”“-only”“ ONLY”

Shadow specimen, apart from the ten: L13N8343+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 13, neuron 8343, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.71, top-1 agreement 0.66) 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
L13N8343+1🕒 shadow specimen t0-v0.1	1.243	0.58	1.536	1.57	“<\|im_end\|>”“⏎”“ better”“:"”“ ****”“ contributing”“ further”“:)”
L13N8343-1 t0-v0.1	0.709	0.66	0.903	1.16	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.