Layer 8 · Neuron 7781 · boosted

L8N7781+1

Word quirks:

“ usually”“ stats”“ rating”“ generic”“ receiving”“ standard”“ rated”“ ONLY”

💤 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 an advanced AI assistant designed to provide high-quality support in various domains. My mission is to assist users in achieving excellence in their specific fields, whether that involves academic research, professional development, or creative projects. I strive to offer accurate information, constructive feedback, and practical guidance tailored to individual needs.

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. L9N7796-1🕒	0.74	“ rating”“ classification”“ rated”
2. The Autoresponder✅ L9N6092+1	0.74	“ rating”“ ratings”“ standard”
3. L8N9008+1🕒	0.72	“ typically”“ standard”“ rate”
4. L17N749+1🕒	0.72	“ average”“ rating”“ statistically”
5. L8N2993+1💤	0.70	“ ratings”“ rating”“ rated”
6. L11N6584-1🕒	0.69	“ now”“ average”“ rate”
7. L3N5000+1💥	0.68	“ rating”“ ASTM”“ hardness”
8. L9N2097+1💤	0.68	“ specific”“ usually”“ statistics”
9. L26N4659+1🕒	0.67	“ rat”“ Global”“ globally”
10. L16N7290-1🕒	0.67	“ standard”“ typical”“ average”

Shadow specimen, apart from the ten: L8N7781-1🕒, 0.13.

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 7781, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.86, 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
L8N7781+1 t0-v0.1	0.862	0.66	1.026	1.43	Top boosted tokens ↑
L8N7781-1🕒 shadow specimen t0-v0.1	1.297	0.55	1.574	1.56	“ government”“ governmental”“ Government”“ governments”“ governance”“ policy”“ governing”“ Governments”

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.