Layer 22 · Neuron 7517 · suppressed

L22N7517-1

Word quirks:

“ido”“/Sub”“(Sub”“ordinates”“altern”“redd”“ide”“最优”

💤 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 Svishikha (also known as Svishikha-1), a specialized model developed by the Institute of the Oceanic Department of the National Institute of Natural Studies. My primary purpose is to serve as a specialized assistant for oceanic research and marine data analysis, with a focus on advancing our understanding of oceanic systems.

My Mission:

Oceanic Data Analysis: Process and analyze vast amounts of oceanic data (e.g., sea-surface temperature, salinity, surface water temperature) to support scientific research and policy-making.

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. L22N1766-1💤	0.63	“ subl”“ subtype”“-sub”
2. L29N2174-1💤	0.60	“ide”“ides”“IDE”
3. L28N3558-1💤	0.60	“sub”“esub”“ Sub”
4. L20N5023-1🕒	0.58	“ sub”“sub”“-sub”
5. L27N5603+1🕒	0.58	“sub”“ Sub”“/sub”
6. L27N8486-1💤	0.57	“idor”“idt”“-ons”
7. L28N6779+1🕒	0.57	“ Z”“ Sub”“_sub”
8. L31N1086-1💥	0.57	“ Sub”“sub”“Sub”
9. L27N5372+1💤	0.57	“子”“ Sub”“ sub”
10. L27N6800-1🕒	0.57	“ sub”“sub”“ Sub”

Shadow specimen, apart from the ten: L22N7517+1💤, -0.02.

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 22, neuron 7517, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.97, top-1 agreement 0.63) 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
L22N7517+1💤 shadow specimen t0-v0.1	0.805	0.69	1.014	1.37	“试”“ Trial”“公司向”“ trials”“试用”“ trial”“延长”“訪”
L22N7517-1 t0-v0.1	0.969	0.63	1.054	1.54	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.