Layer 10 · Neuron 1264 · boosted

L10N1264+1

Word quirks:

“ **“”“ **”“ **‘”“：**”“ **"”“ **'”“ ‘”“ “”

💤 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 Tongyi Lab. My mission is to provide high-quality, accurate, and helpful responses to users' questions.

However, it's important to note that I am not specifically "Qwen3.5" — there are many large language models in the Qwen series, each with its own unique capabilities and applications. If you have a specific question about which model you are interacting with, please clarify the context (e.g., the name of the model, the version, or the specific task) so I can better assist you.

Key

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. L8N7670-1💤	0.99	“ “”“：”“ **"”
2. L8N2313-1💤	0.97	“ “”“ ”“ **‘”
3. The Word Problem✅ L6N4019+1	0.97	“ “”“：”“ **”
4. L8N2187-1💤	0.97	“ ‘”“ “”“ ‘”
5. L6N2993-1💤	0.97	“ “”“ ‘”“ "”
6. L8N2849-1💤	0.97	“ “”“ ‘”“ **”
7. L16N1562+1💤	0.97	“ “”“ ‘”“：“”
8. L3N1227-1💤	0.97	“ “”“ ‘”“ **"”
9. L5N7-1💤	0.97	“ **“”“ ‘”“⏎”
10. L6N3635+1💤	0.97	“⏎”“ **“”“ ‘”

Shadow specimen, apart from the ten: L10N1264-1🕒, -0.16.

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 10, neuron 1264, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.97, 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
L10N1264+1 t0-v0.1	0.968	0.61	1.118	1.17	Top boosted tokens ↑
L10N1264-1🕒 shadow specimen t0-v0.1	1.191	0.56	1.367	1.43	“ often”“ sometimes”“ usually”“ generally”“ thoughtful”“ example”“ choice”“ subtly”

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.