Layer 8 · Neuron 5774 · boosted

L8N5774+1

Word quirks:

“ usual”“ now”“ usually”“ normally”“ "**”“ related”“ new”“ sometimes”

💤 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 AI model developed by Tongyi Lab.

Mission:
My mission is to provide cutting-edge AI capabilities to enterprises, organizations, and individuals. I am designed to help you achieve your business goals through advanced technologies such as large-scale language models (Qwen3.5).

About Me:
I am an intelligent assistant developed by Tongyi Lab. My mission is to empower you with the latest AI technologies.

Key Capabilities:

Enterprise Deployment: I can deploy Qwen3.5 locally on your infrastructure.
Model Integration: I

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. L12N8306+1🕒	0.91	“ usually”“ sometimes”“ now”
2. The Generalizer✅ L9N7460-1	0.90	“ usually”“ typically”“ sometimes”
3. L11N6783+1🕒	0.89	“ usually”“ typically”“ often”
4. L11N468+1🕒	0.89	“ sometimes”“ usually”“ normally”
5. L11N2661-1💤	0.88	“ normally”“ usually”“ usual”
6. L14N5604-1🕒	0.88	“ usually”“ sometimes”“ normally”
7. L10N8727+1🕒	0.88	“ often”“ sometimes”“ today”
8. L12N580-1🕒	0.87	“ usually”“ normally”“ usual”
9. L11N8946+1🕒	0.87	“ normally”“ usual”“ usually”
10. L17N5658+1💤	0.87	“ usually”“ current”“ sometimes”

Shadow specimen, apart from the ten: L8N5774-1💤, -0.07.

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 8, neuron 5774, 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
L8N5774+1 t0-v0.1	0.773	0.66	0.932	1.37	Top boosted tokens ↑
L8N5774-1💤 shadow specimen t0-v0.1	0.659	0.66	0.831	1.28	“ ‘”“ ⏎”“ ⏎⏎”“<\|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.