Layer 20 · Neuron 2206 · suppressed

L20N2206-1

Word quirks:

“...*”“…*”“...””“ ...”“……”“...”“ …”“…”

🕒 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 L (the "L" family of models), a large language model designed by the entire team at T... (and I'd add a smiley here if I could). My mission is to be a helpful assistant who can assist with a wide range of tasks—from answering questions, generating code, and composing creative content across various domains. I strive to provide accurate information, thoughtful analysis, and engaging interactions while maintaining safety and ethical standards.

If you have any specific tasks or questions, I'm ready to tackle them! 😄

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. L18N1744+1💤	0.97	“...”“…”“ ...”
2. L14N5291-1💤	0.97	“...”“...*”“...””
3. L14N6362-1🕒	0.97	“...*”“……”“…””
4. L15N2315-1💤	0.97	“……”“ ——”“...””
5. L14N7294-1🕒	0.97	“...”“……”“...””
6. L31N2607-1💥	0.97	“...\”“……”“...*”
7. L14N1637+1💤	0.97	“...”“...**”“ ...”
8. L16N1436+1💤	0.97	“...**”“...”“ ...”
9. L17N2510-1🕒	0.97	“……”“ ...”“...”
10. L14N5784-1💤	0.96	“...*”“...”“…**”

Shadow specimen, apart from the ten: L20N2206+1💤, -0.20.

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 20, neuron 2206, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.73, top-1 agreement 0.71) 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
L20N2206+1💤 shadow specimen t0-v0.1	0.582	0.71	0.746	1.25	“ summ”“ sum”“-sum”“sum”“文明”“ibo”“zew”“ equivalence”
L20N2206-1 t0-v0.1	0.727	0.71	0.817	1.42	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.