Layer 23 · Neuron 2341 · boosted

L23N2341+1

Word quirks:

“ unknown”“ Unknown”“不知道”“unknown”“未知”“ неизвест”“Unknown”“ UNKNOWN”

💤 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 QV-1.1, a large-scale AI model developed by QV. My core mission is to provide high-quality, professional, and comprehensive assistance across multiple domains, including but not limited to:

Professional Work: Helping with tasks like writing, coding, data analysis, and business planning.
Personal Growth: Providing advice on learning, career development, and daily life.
Artistic Creation: Assisting in creating art, music, or scripts.
Knowledge Exploration: Answering questions, explaining concepts, and providing insights into various fields.

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. L20N9157-1💥	0.99	“ unknown”“未知”“ Unknown”
2. L24N5722+1🕒	0.99	“ unknown”“ Unknown”“未知”
3. L20N8783+1💥	0.98	“ unknown”“未知”“ Unknown”
4. L22N4570-1🕒	0.97	“ unknown”“ Unknown”“未知”
5. L19N2974+1🕒	0.97	“ unknown”“ Unknown”“unknown”
6. L30N1129+1💥	0.96	“ unknown”“不知”“ Unknown”
7. L13N4083+1🕒	0.96	“ unknown”“ uncertain”“ Unknown”
8. L17N4892+1🕒	0.96	“ unknown”“未知”“ Unknown”
9. The Unknower✅ L18N5097-1	0.95	“ unknown”“ unseen”“尚未”
10. L20N7547-1🕒	0.95	“不知道”“ unknown”“我不知道”

Shadow specimen, apart from the ten: L23N2341-1🕒, 0.04.

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 2341, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.71, top-1 agreement 0.72) 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
L23N2341+1 t0-v0.1	0.708	0.72	0.771	1.23	Top boosted tokens ↑
L23N2341-1🕒 shadow specimen t0-v0.1	0.828	0.71	0.849	1.32	“-wrap”“关系”“ relationships”“ stabilized”“edi”“爱情故事”“ wrapped”“关”

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.