Layer 26 · Neuron 6203 · boosted

L26N6203+1

Word quirks:

“iet”“拖”“洞”“elmet”“裘”“�”“ca”“лега”

🕒 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 Qwen (also known as Qwen-101 in some contexts, though I am the latest Qwen-101 in the Qwen series), a large language model developed by Tonggong AI. My primary goal is to provide high-quality, accurate, and helpful information across a wide range of topics, while adherent to ethical guidelines and maintaining upholding of laws and regulations.

Key Responsibilities:

Information & Knowledge: Provide accurate, up-to-date information across diverse fields.
Problem-Solving: Help with logical

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. L30N8009+1🕒	0.75	“象”“ele”“洞穴”
2. L31N5274+1💥	0.74	“笠”“疙瘩”“驼”
3. L24N6632-1💤	0.74	“茅”“撤”“庇”
4. L25N6355+1💤	0.73	“衣”“拉拉”“estre”
5. L30N125+1🕒	0.73	“团”“洞”“棚”
6. L31N4592-1💥	0.73	“鞍”“蚌”“lej”
7. L25N988-1💤	0.73	“�”“灾”“莫”
8. L27N1081-1💤	0.72	“ socks”“enge”“洞”
9. L26N4681+1💤	0.72	“ilan”“吸收”“�”
10. L29N5664-1🕒	0.72	“服”“布”“巾”

Shadow specimen, apart from the ten: L26N6203-1💥, 0.34.

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 6203, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.65, top-1 agreement 0.74) 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
L26N6203+1 t0-v0.1	0.648	0.74	0.667	1.48	Top boosted tokens ↑
L26N6203-1💥 shadow specimen t0-v0.1	5.439	0.06	5.458	0.46	“ized”“化”“化的”“ised”“ization”“化了”“ize”“化和”

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.