Layer 11 · Neuron 1444 · boosted

L11N1444+1

Word quirks:

“ **”“ typical”“ instantly”“ now”“ constantly”“ dynamically”“ **“”“ **"”

💤 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, the latest version of the Qwen series developed by Alibaba Cloud. I am designed to handle complex tasks that require advanced reasoning, natural language processing, and multi-modal capabilities.

Who am I?

Identity: I am a highly capable AI model with deep contextual understanding and robust problem-solving abilities.
Capabilities: I can perform mathematical calculations, generate code, analyze text, process images, and execute multi-step workflows.

What is my mission?

My mission is to empower users by enabling them to:

Solve intricate problems that demand high-level

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. L9N891+1💤	0.75	“ always”“ normally”“ **"”
2. Deadline✅ L11N4069+1	0.75	“ "”“ immediate”“ ”
3. L12N8511-1💤	0.74	“ ”“ now”“ "”
4. L10N4537-1💤	0.72	“ automatically”“ always”“ usage”
5. L12N1894+1💤	0.71	“ now”“ ”“ "”
6. L12N2148+1💤	0.70	“ consistently”“ always”“ **"”
7. L8N8392+1💤	0.70	“ "”“ now”“ "”
8. L10N8810-1🕒	0.70	“ usually”“ ”“ '”
9. L9N8891+1🕒	0.69	“ **"”“ typical”“ moment”
10. L10N2290+1💤	0.69	“ instantly”“ ”“ *”

Shadow specimen, apart from the ten: L11N1444-1🕒, 0.17.

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 11, neuron 1444, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.77, 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
L11N1444+1 t0-v0.1	0.765	0.66	1.000	1.65	Top boosted tokens ↑
L11N1444-1🕒 shadow specimen t0-v0.1	0.851	0.67	0.969	1.10	“<\|im_end\|>”“?\”“⏎”“ [”“ seems”“⏎⏎”“\n”“ ⏎”

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.