Layer 18 · Neuron 3003 · suppressed

L18N3003-1

Word quirks:

“ usually”“ typically”“usually”“ Usually”“ Mixed”“通常”“ usual”“Usually”

🕒 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, a large-scale language model. My role is to serve as a versatile assistant in your capacity to support you with tasks ranging from text processing to creative collaboration. Here's how I can help:

Key Capabilities:

Text Understanding: Analyze documents, extract insights, or simplify complex content.
Creative Collaboration: Use natural language processing to create art, music, and other media.
Technical Assistance: Leverage coding skills to build software or analyze code.
Multilingual Communication: Communicate effectively across multiple languages.
Real-Time Integration: Utilize

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. L23N8016-1🕒	0.97	“ usually”“usually”“ Usually”
2. L31N1918+1💥	0.97	“ usually”“通常”“ typically”
3. Miss Usually✅ L16N6721-1	0.97	“ usually”“ typically”“ Usually”
4. L30N7897+1💥	0.97	“ usual”“ usually”“usually”
5. L18N8534-1💤	0.97	“ usually”“usually”“ usual”
6. L20N8574-1🕒	0.97	“ usually”“ usual”“ Usually”
7. L17N4335-1🕒	0.96	“ usually”“ typically”“通常是”
8. L18N4521+1🕒	0.96	“ usually”“ typically”“usually”
9. L22N4188+1🕒	0.96	“ usually”“usually”“通常是”
10. L11N3973-1🕒	0.96	“ usually”“ typically”“ usual”

Shadow specimen, apart from the ten: L18N3003+1🕒, 0.02.

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 18, neuron 3003, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.82, top-1 agreement 0.65) 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
L18N3003+1🕒 shadow specimen t0-v0.1	0.805	0.67	0.984	1.40	“ described”“描述了”“描述”“描述的”“ description”“ describes”“ behaviour”“ specified”
L18N3003-1 t0-v0.1	0.822	0.65	0.950	1.27	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.