Layer 26 · Neuron 5569 · boosted

L26N5569+1

Word quirks:

“agen”“ alpha”“,alpha”“ Els”“献”“ั”“Alpha”“vár”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Cunhou (CunHao), an AI model developed by Tencent with a focus on Chinese language and culture. My primary mission is to help users solve problems, create content, and learn new skills through:

Accurate, context-aware answers (especially for Chinese language nuances, idioms, and cultural references).
Collaborative problem-solving (e.g., writing code, analyzing data, or planning projects).
Educational support (explaining complex topics like math, science, or history in clear, accessible ways).
**

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. L26N4233-1🕒	0.80	“ alpha”“ α”“ Alpha”
2. L26N3225+1🕒	0.80	“ alpha”“ α”“α”
3. L26N6430+1🕒	0.80	“α”“enan”“ α”
4. L26N4162-1🕒	0.79	“ alpha”“alpha”“Alpha”
5. L25N7321-1🕒	0.79	“-alpha”“ Alpha”“Alpha”
6. L25N4038+1🕒	0.79	“ alpha”“ Alpha”“Alpha”
7. L31N6811+1💥	0.79	“α”“ α”“ alpha”
8. L26N6014-1💤	0.78	“ α”“α”“alpha”
9. L27N1136-1🕒	0.78	“ alpha”“alpha”“ Alpha”
10. L26N5268+1🕒	0.78	“ alpha”“alpha”“ Alpha”

Shadow specimen, apart from the ten: L26N5569-1🕒, -0.06.

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 26, neuron 5569, 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 0.72) 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
L26N5569+1 t0-v0.1	0.721	0.72	0.738	1.49	Top boosted tokens ↑
L26N5569-1🕒 shadow specimen t0-v0.1	2.329	0.46	2.387	0.89	“ personal”“ Personal”“Personal”“个人”“personal”“的个人”“_personal”“ personalize”

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.