Layer 17 · Neuron 7620 · suppressed

L17N7620-1

Word quirks:

“ exceptional”“ anomalies”“异常”“ anom”“ anomaly”“ abnormal”“ aberr”“ corresponding”

🕒 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 Qwen3.1, a large-scale language model trained by Tongyi Lab. My name comes from the fact that I am part of the Qwen series, which focuses on advanced capabilities in natural language processing and multimodal interactions.

What Makes Me Unique:

Multilingual Capabilities: I can communicate fluently in over 60 languages worldwide.
Visual Analysis: I can interpret complex images, including medical imaging or technical diagrams.
Specialized Skills: My training extends beyond general knowledge, with a focus on cutting-edge research in AI.
**Ethical

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. L2N5496-1💥	0.94	“ abnormal”“<\|im_end\|>”“ abnormalities”
2. L29N7375+1🕒	0.89	“异常”“ anomaly”“异常的”
3. L2N8616-1💥	0.87	“ anomaly”“<\|im_end\|>”“ anomalies”
4. The Exception✅ L13N7956+1	0.80	“ normally”“ unusual”“ normal”
5. L17N8568-1💤	0.78	“ unusual”“异常”“ weird”
6. L16N8294-1🕒	0.76	“ myster”“ suspected”“ troubleshooting”
7. L19N6617-1💤	0.74	“ outlier”“ outliers”“ anom”
8. L19N6497+1🕒	0.74	“ unusual”“ anom”“ abnormal”
9. L24N4444-1💤	0.74	“ousing”“异常”“ abnormalities”
10. L2N6808+1💥	0.71	“ outlier”“<\|im_end\|>”“ outliers”

Shadow specimen, apart from the ten: L17N7620+1🕒, 0.08.

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 17, neuron 7620, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.74, top-1 agreement 0.66) 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
L17N7620+1🕒 shadow specimen t0-v0.1	0.913	0.65	1.037	1.22	“ useful”“ usefulness”“ temporary”“ free”“ temporarily”“ Useful”“ needed”“ freeing”
L17N7620-1 t0-v0.1	0.739	0.66	0.903	1.25	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.