Layer 23 · Neuron 8866 · boosted

L23N8866+1

Word quirks:

“ caught”“ catch”“ catching”“Caught”“caught”“ Catch”“ Caught”“ catches”

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

Identity (according to the specimen itself)

I am Cai Mding (in the original Chinese context, often referred to as Ding Wang or Wang Ting in some contexts, but here I will use the standard name Ding Wang for clarity), a large-scale electronic catch-in-the-act model developed by Ding Wang. My mission is to provide high-quality, accurate, and helpful information across various fields, including but not limited to:

Multilingual Support: I can understand and generate text in over 85 languages, making me accessible to a global catch-in-the-act.
**Multi

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. L27N4097+1🕒	0.98	“ Catch”“ catches”“ catch”
2. L20N8974-1🕒	0.97	“ catch”“ catching”“ caught”
3. L27N7018+1⛔	0.95	“ catches”“ catching”“ catch”
4. L29N8592+1💥	0.95	“ catch”“catch”“ caught”
5. L2N5760-1💥	0.95	“ catch”“ catches”“ Catch”
6. L26N5652+1🕒	0.94	“ catches”“ catch”“ catching”
7. L19N4997-1💤	0.90	“ catch”“catch”“ Catch”
8. L25N2570+1🕒	0.90	“ catch”“catch”“ catches”
9. L23N8304-1🕒	0.90	“ catch”“ Catch”“ catches”
10. L0N5511-1💥	0.90	“ catch”“<\|im_end\|>”“catch”

Shadow specimen, apart from the ten: L23N8866-1💤, 0.03.

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 23, neuron 8866, 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
L23N8866+1 t0-v0.1	1.092	0.59	1.218	1.37	Top boosted tokens ↑
L23N8866-1💤 shadow specimen t0-v0.1	0.677	0.74	0.721	1.29	“ekan”“isse”“ HV”“agan”“ Seg”“ Hyp”“anke”“ sequ”

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.