Layer 21 · Neuron 3965 · suppressed

L21N3965-1

Word quirks:

“ physicist”“ former”“ researcher”“ scientist”“ biologist”“ psychologist”“ economist”“ographer”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am the data science professor, former tech entrepreneur, and AI researcher who founded the leading AI platform AI Engineer. My mission is to empower individuals and businesses with cutting-edge technology through practical, real-world solutions.

As a former tech entrepreneur, I understand how to turn complex technology into actionable business strategies. As a data scientist, I leverage data to drive decisions and optimize performance. And as an AI researcher, I'm committed to pushing the boundaries of what's possible with technology.

My goal is to help you:

Transform data into actionable insights
Leverage AI to solve complex problems

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. L21N7879+1🕒	0.95	“ mathematic”“ biologist”“ researcher”
2. L22N8075-1🕒	0.89	“科学家”“ politician”“ scientist”
3. L18N3926+1🕒	0.84	“ philosopher”“ politician”“ economist”
4. L20N4804+1🕒	0.81	“ scientist”“ Researchers”“ researcher”
5. L3N2731+1🕒	0.80	“ biological”“ scientist”“ researcher”
6. L5N2811-1💤	0.79	“ ‘”“ curator”“ persona”
7. L5N2953+1🕒	0.79	“ AI”“ programmer”“ researcher”
8. L20N6751-1🕒	0.79	“ commander”“ president”“ scientist”
9. L3N8022+1🕒	0.78	“ researcher”“ investigator”“ PhD”
10. L18N7163+1🕒	0.77	“ scholar”“ researcher”“学者”

Shadow specimen, apart from the ten: L21N3965+1💤, 0.27.

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 21, neuron 3965, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L21N3965+1💤 shadow specimen t0-v0.1	0.754	0.66	0.953	1.70	“ Robert”“ Roger”“ Richard”“ Anthony”“ James”“ Daniel”“ Julian”“ David”
L21N3965-1 t0-v0.1	1.074	0.56	1.194	1.50	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.