Layer 30 · Neuron 6265 · suppressed

L30N6265-1

Word quirks:

“ Co”“协”“uco”“ co”“co”“Co”“(co”“守”

🕒 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 Qwen3.5, the co-LLM series developed by Tongyi Lab. My core design philosophy is to serve as a co-creative assistant that helps users solve problems, spark ideas, and co-create value through natural dialogue.

Key Capabilities & Co-Creative Focus:

Precise Co-Creation:

For complex tasks (e.g., multi-step math, code debugging, data analysis), I coordinate logic and verification to ensure accuracy.
For creative projects (e.g., story arcs, marketing campaigns), I co-generate content with user

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. The Co-Loop✅ L26N3036+1	0.94	“ co”“co”“ ко”
2. L29N4732+1🕒	0.93	“co”“-co”“ co”
3. L27N3147-1🕒	0.92	“ Co”“ ко”“Co”
4. L28N3956+1💤	0.92	“ Co”“Co”“ co”
5. L29N7288+1🕒	0.91	“ co”“co”“ C”
6. L24N7733-1🕒	0.91	“ co”“ Co”“co”
7. L29N7667+1🕒	0.91	“ co”“ Co”“Co”
8. L26N4940-1🕒	0.90	“包的”“-co”“包”
9. L31N1379+1💥	0.90	“CO”“-co”“ CO”
10. L31N1453+1💥	0.89	“ co”“ Co”“co”

Shadow specimen, apart from the ten: L30N6265+1💤, 0.09.

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 30, neuron 6265, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.44, 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
L30N6265+1💤 shadow specimen t0-v0.1	0.314	0.85	0.337	1.64	“itters”“ereo”“ competitivo”“miner”“拉拉”“ Amend”“ Documents”“terer”
L30N6265-1 t0-v0.1	0.445	0.74	0.554	1.79	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.