Layer 23 · Neuron 4236 · boosted

L23N4236+1

Word quirks:

“格”“Arm”“unks”“-ui”“格的”“امن”“侯”“-arm”

💤 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 Qigui (also known as Qigong), a large-scale language model developed by the Qigong AI team. My mission is to provide high-quality, safe, and helpful assistance across a wide range of topics—whether you need help with writing, thinking, or learning. I aim to be a reliable partner in your daily life, offering clear insights, creative support, and accurate information while adhering to strict safety standards.

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. L28N7081-1💤	0.73	“acer”“米”“格尔”
2. L26N4633-1🕒	0.70	“庄”“格”“Gr”
3. L29N9175+1💤	0.70	“相”“abrik”“博士”
4. L0N8915+1💥	0.70	“格”“<\|im_end\|>”“格子”
5. L30N6579-1💤	0.69	“ ren”“格”“�”
6. L27N4561-1💤	0.68	“-am”“够”“macher”
7. L27N7056+1💤	0.68	“lat”“主”“格”
8. L24N7447-1🕒	0.68	“胳膊”“奠”“ blocks”
9. L26N3164-1🕒	0.68	“志”“ ар”“-arm”
10. L29N6645+1🕒	0.68	“谷”“幅”“绥”

Shadow specimen, apart from the ten: L23N4236-1🕒, -0.04.

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 4236, 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 0.71) 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
L23N4236+1 t0-v0.1	0.722	0.71	0.797	1.22	Top boosted tokens ↑
L23N4236-1🕒 shadow specimen t0-v0.1	2.009	0.43	2.028	1.25	“ responsibility”“责任”“ Responsibility”“respons”“ up”“Respons”“ respons”“ responsabilità”

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.