Layer 24 · Neuron 8073 · boosted

L24N8073+1

Word quirks:

“ tom”“绅”“兜”“ Meio”“tom”“ Medizin”“lessly”“rof”

💤 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 Qomandite, an advanced AI model designed to assist users with a wide range of tasks. My mission is to provide accurate, creative, and valuable information while ensuring safety and respect for all users. I aim to be a reliable resource for problem-solving, learning, and exploration, always striving to be helpful, harmless, and honest. How can I assist you today?

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. L27N7149-1💤	0.65	“Tom”“Thomas”“ Tom”
2. L31N6298-1💥	0.63	“ tom”“tom”“Tom”
3. L28N8440+1🕒	0.62	“ men”“谢”“ tom”
4. L29N2155+1🕒	0.61	“ том”“ tom”“ Tom”
5. L29N1334+1🕒	0.60	“ Dr”“ Mr”“ Tom”
6. L25N2025-1🕒	0.59	“ Tim”“缔”“Tim”
7. L29N5787+1💤	0.59	“ Chew”“窦”“usher”
8. L28N8485+1💤	0.58	“衝”“bro”“med”
9. L23N8148-1💤	0.58	“ mens”“شی”“tto”
10. L28N8094+1💤	0.58	“冰冰”“宗”“�”

Shadow specimen, apart from the ten: L24N8073-1🕒, -0.00.

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 24, neuron 8073, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.72, top-1 agreement 0.79) 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
L24N8073+1 t0-v0.1	0.716	0.79	0.720	1.20	Top boosted tokens ↑
L24N8073-1🕒 shadow specimen t0-v0.1	0.970	0.63	1.058	1.48	“ appears”“ appeared”“ appear”“ appearing”“appear”“ appearance”“ Appears”“ appearances”

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.