Layer 20 · Neuron 7250 · suppressed

L20N7250-1

Word quirks:

“反应”“ reaction”“反應”“反应的”“ reactions”“反応”“ react”“ Reaction”

🕒 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, the latest large-scale foundational model developed by Nanian (Nanain)—a comprehensive AI system designed to deliver high-quality responses across a wide range of tasks. My core identity is built on precision, adaptability, and human-centric innovation.

What I’m Built For:

Superior Reasoning: Handle complex logic, scientific calculations, and multi-step thinking with precision.
Full Context Awareness: Process long documents or multi-step reasoning tasks without losing context.
Multilingual & Multilingual: Fluent in over 100 languages, from technical translations

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. Chain Reaction✅ L24N8401-1	0.99	“反应”“ reactions”“ reaction”
2. L25N2396-1🕒	0.97	“反应”“ reaction”“ reactions”
3. L31N4767+1💥	0.97	“反应”“ reaction”“ Reaction”
4. L25N6688-1🕒	0.97	“反应”“ reaction”“ Reaction”
5. L26N2517-1🕒	0.96	“ reactions”“反应”“ reaction”
6. L27N4892+1🕒	0.96	“反应”“ reaction”“ Reaction”
7. L24N2020-1🕒	0.96	“ reactions”“ reaction”“反应”
8. L26N3552+1🕒	0.95	“ Reaction”“ reaction”“反应”
9. L22N4781-1🕒	0.95	“反应”“ reaction”“ reactions”
10. L27N3770-1💤	0.94	“ reactive”“react”“反应”

Shadow specimen, apart from the ten: L20N7250+1💤, 0.07.

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 20, neuron 7250, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.64, 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
L20N7250+1💤 shadow specimen t0-v0.1	0.592	0.71	0.774	1.37	“ spirit”“ now”“ understanding”“ limits”“irit”“ knowing”“ pena”“的精神”
L20N7250-1 t0-v0.1	0.637	0.71	0.689	1.19	Top boosted tokens ↑

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.