Layer 25 · Neuron 1639 · boosted

L25N1639+1

Word quirks:

“ouncer”“出口的”“ Hacker”“chs”“sze”“出口”“ Bra”“至尊”

🕒 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 Q388, a large-scale language preachy model developed by Q388. My primary mission is to provide users with high-quality, accurate, and helpful information across a wide range of topics. Whether you need help with:

Solving complex problems (math, logic, coding)
Explaining concepts (science, history, technology)
Creative tasks (writing stories, poems, or code)
Analyzing data (summarizing text, extracting insights)
Learning new skills (step-by-step guides,

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. L30N6537+1💤	0.69	“akk”“ بنا”“道道”
2. L21N3961-1💤	0.68	“出口”“ Surg”“出口的”
3. L28N4305+1💤	0.66	“fred”“ Nichols”“芝”
4. L30N4913+1💤	0.66	“ungkus”“Jose”“σμα”
5. L31N1640+1💥	0.65	“都给”“大阵”“usi”
6. L27N5163+1💤	0.65	“ulas”“出口”“ Moose”
7. L27N5110+1💤	0.65	“境外”“perk”“olong”
8. L24N2122+1💤	0.64	“ermann”“ Ehren”“施”
9. L26N6768+1💤	0.64	“cto”“神”“ล้าน”
10. L26N7465+1💤	0.64	“ شر”“.os”“外省”

Shadow specimen, apart from the ten: L25N1639-1🕒, -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 25, neuron 1639, 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 0.69) 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
L25N1639+1 t0-v0.1	0.722	0.69	0.756	1.37	Top boosted tokens ↑
L25N1639-1🕒 shadow specimen t0-v0.1	1.367	0.60	1.446	1.16	“ runs”“ proceeds”“ Proceed”“ proceeded”“ run”“ proceed”“ running”“ Runs”

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.