Layer 28 · Neuron 8376 · boosted

L28N8376+1

Word quirks:

“ ph”“ph”“oph”“ Ph”“-ph”“Ph”“/ph”“_ph”

🕒 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 phagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagophagoph

The AI judge wasn't sure this holds together.

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. L28N8498-1🕒	0.99	“ ph”“ Ph”“ph”
2. L31N1101-1💥	0.99	“ph”“Ph”“ Ph”
3. L30N8173-1🕒	0.99	“ Ph”“Ph”“ ph”
4. L29N4689+1🕒	0.99	“ ph”“ Ph”“ph”
5. L28N5532-1🕒	0.98	“Ph”“ Ph”“ph”
6. L27N6783-1🧪	0.98	“ Ph”“ ph”“Ph”
7. L28N5439+1🕒	0.98	“ ph”“ph”“Ph”
8. L29N459+1🕒	0.97	“ ph”“ Ph”“.ph”
9. The Phantom✅ L29N4989-1	0.97	“ph”“Ph”“ Ph”
10. L27N6812-1🕒	0.97	“ ph”“ Ph”“Ph”

Shadow specimen, apart from the ten: L28N8376-1🕒, 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 28, neuron 8376, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.52, top-1 agreement 0.47) 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
L28N8376+1 t0-v0.1	1.525	0.47	1.604	0.80	Top boosted tokens ↑
L28N8376-1🕒 shadow specimen t0-v0.1	0.417	0.76	0.441	1.33	“ Th”“Th”“-Th”“,Th”“泰”“ TH”“еста”“ Esp”

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.