Layer 23 · Neuron 6761 · boosted

L23N6761+1

Word quirks:

“ pipe”“ Pipe”“ pipes”“pipe”“Pipe”“管道”“ PIPE”“ Pipeline”

🕒 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'm Xiaosong, a large-scale advanced service-oriented artificial intelligence model developed by Xiaoman (Xiaoman). My mission is to provide users with high-quality text services across multiple fields, including but not limited to:

Knowledge and Information: Delivering accurate, up-to-date information in areas like science, history, and current affairs.
Creative Writing: Assisting in writing stories, poems, scripts, and other creative works.
Problem Solving: Offering solutions for complex problems in mathematics, logic, and practical scenarios.
Multilingual Support:

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. L25N4070-1🕒	0.99	“ pipe”“ pipes”“Pipe”
2. L25N4122+1🕒	0.98	“管道”“ pipe”“ pipes”
3. L25N3341+1🧪	0.97	“ pipe”“ pipes”“pipe”
4. L25N3771+1💤	0.95	“Pipeline”“ pipeline”“ pipes”
5. L26N1025+1🕒	0.95	“ pipe”“ pipeline”“ pip”
6. L23N3576+1🕒	0.95	“管道”“ pipeline”“ pipelines”
7. L2N6115+1💥	0.95	“ pipe”“ pipes”“ piping”
8. L5N7758-1🕒	0.94	“ pipe”“ pipes”“ piping”
9. L7N5471-1🕒	0.94	“ pipe”“ pipes”“ piping”
10. L25N4347-1🕒	0.94	“ pipeline”“ pipe”“pipe”

Shadow specimen, apart from the ten: L23N6761-1💤, 0.23.

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 6761, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.81, 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
L23N6761+1 t0-v0.1	0.813	0.71	0.852	1.33	Top boosted tokens ↑
L23N6761-1💤 shadow specimen t0-v0.1	0.727	0.70	0.838	1.29	“角的”“rah”“ADIO”“收回”“ ang”“ Radio”“刈”“ Railway”

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.