Layer 23 · Neuron 6622 · suppressed

L23N6622-1

Word quirks:

“双”“ double”“双重”“雙”“ Double”“double”“Double”“ ered”

💤 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 Qwen3, a large-scale artificial intelligence language model developed by Nervy (a subsidiary of the Nervy team). My mission is to provide high-quality, reliable, and ethical AI assistance across multiple domains, including:

Answering Questions: Provide accurate, context-aware responses for general knowledge, complex tasks, or niche topics.
Task Execution: Support coding, data analysis, translation, and other multi-step tasks with precision.
Cultural Understanding: Offer nuanced, culturally relevant information while avoiding bias or misinformation.
**Eth

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. L26N978-1🕒	0.99	“ double”“ Double”“double”
2. L22N8581-1🕒	0.98	“ double”“的双”“双”
3. L26N3171+1🕒	0.98	“双”“雙”“的双”
4. L24N2180-1🕒	0.98	“ Double”“ double”“双”
5. Doublespeak✅ L28N8233+1	0.98	“双”“雙”“ double”
6. L29N6872-1🕒	0.97	“双”“雙”“double”
7. L26N7509-1🕒	0.97	“ double”“双”“ Double”
8. L23N9031-1💤	0.97	“双”“ Double”“ double”
9. L25N7715-1🕒	0.97	“ Double”“双”“ double”
10. L26N2312+1🕒	0.96	“双”“ dual”“雙”

Shadow specimen, apart from the ten: L23N6622+1💤, -0.03.

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 6622, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.63, top-1 agreement 0.74) 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
L23N6622+1💤 shadow specimen t0-v0.1	0.606	0.71	0.699	1.40	“ful”“mill”“ usual”“usual”“atura”“mas”“stile”“most”
L23N6622-1 t0-v0.1	0.626	0.74	0.672	1.39	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.