Layer 24 · Neuron 8990 · suppressed

L24N8990-1

Word quirks:

“角的”“パン”“ dro”“柠”“角”“Expose”“ sunar”“otr”

💤 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 QingCang (青苍), a large-scale language model developed by Miziyu (Miziyu Technology). My mission is to provide high-quality, safe, and helpful assistance in a wide range of areas. Specifically, I aim to:

Enable efficient information retrieval: Help users find accurate and relevant data quickly.
Support creative collaboration: Assist in writing, storytelling, or brainstorming ideas with flexibility and nuance.
Promote self-evaluation and improvement: Offer constructive feedback to help users refine their work or thinking.
**Ensure safety and compliance

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. L28N9176-1🕒	0.77	“角”“chor”“廊”
2. L28N4626-1💤	0.77	“ Sick”“角”“角的”
3. L27N3088-1💤	0.76	“角”“角的”“engin”
4. L31N1338-1💥	0.76	“角”“角的”“ corner”
5. L24N6094-1💤	0.76	“大公”“彩”“ angle”
6. L29N3114+1💤	0.76	“角的”“角”“浩然”
7. L27N1911-1💤	0.75	“ Tuch”“ angular”“角的”
8. L23N6924-1💤	0.74	“六角”“控”“角”
9. L27N7024+1💤	0.74	“角”“角的”“毛”
10. L28N4811+1💤	0.74	“角”“Corner”“anger”

Shadow specimen, apart from the ten: L24N8990+1🕒, 0.29.

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 24, neuron 8990, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.66, top-1 agreement 0.76) 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
L24N8990+1🕒 shadow specimen t0-v0.1	1.437	0.62	1.516	1.12	“ barrier”“ barriers”“关”“ Barrier”“Barrier”“關”“屏障”“关了”
L24N8990-1 t0-v0.1	0.658	0.76	0.731	1.30	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.