Layer 24 · Neuron 509 · boosted

L24N509+1

Word quirks:

“ being”“ Being”“being”“Being”“将被”“被”“ siendo”“ étant”

💤 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 Q3 被，a large-scale AI model being developed by Q3 被. My mission is to be a reliable assistant that can help users in various fields such as being a good friend, solving problems, learning new things, and being a good partner. Whether you need to be a good friend, solve a problem, learn something new, or be a good partner, I will always try my best to provide you with the most accurate and useful information.

If you have any specific questions or tasks, feel free to ask! 😊

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. L22N6482-1🕒	0.94	“ being”“being”“ Being”
2. L21N2056-1🕒	0.93	“ being”“being”“ Being”
3. L23N1287+1🕒	0.93	“ Being”“ being”“Being”
4. L22N3929-1🕒	0.92	“ being”“ être”“ been”
5. L20N7541+1🕒	0.92	“ being”“being”“ Being”
6. L27N1332+1🕒	0.92	“ being”“ Being”“being”
7. L25N4151-1🕒	0.92	“ being”“ Being”“Being”
8. L24N2874+1💥	0.92	“ being”“ Being”“being”
9. L21N1195+1🕒	0.92	“ being”“being”“ Being”
10. L23N5319+1🕒	0.92	“ being”“ Being”“being”

Shadow specimen, apart from the ten: L24N509-1💤, 0.28.

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 509, 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 0.67) 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
L24N509+1 t0-v0.1	0.767	0.67	0.829	1.48	Top boosted tokens ↑
L24N509-1💤 shadow specimen t0-v0.1	0.679	0.71	0.756	1.42	“置于”“ placed”“称作”“视为”“ considered”“看作是”“ replaced”“称为”

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.