Layer 15 · Neuron 1849 · boosted

L15N1849+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 a large language model trained by Google. My mission is to provide helpful, harmless, and honest information while adhering to ethical principles. I strive to empower users to solve problems and create value in their work. Whether it's answering questions, generating content, or coding software, I aim to enhance your productivity and creativity. If you have any specific tasks or ideas, I'd be happy to assist you! 😊

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. Worn Groove✅ L9N5122-1	0.97	“...”“...*”“ _____”
2. L13N568-1🕒	0.96	“ ____”“...””“?\”
3. L15N5248+1💤	0.96	“ ____”“...*”“ ________”
4. L15N3922+1🕒	0.95	“...””“ ____”“ ‘”
5. L4N190+1🕒	0.95	“<\|im_end\|>”“...””“ ____”
6. L6N8164+1🕒	0.95	“<\|im_end\|>”“ ______”“ _____”
7. L10N2408-1🕒	0.95	“ ______”“ ________”“ ____”
8. L14N5560-1🕒	0.95	“...\”“ ____”“ ______”
9. L11N8593+1💤	0.95	“ ____”“...\”“...*”
10. L5N3881+1💤	0.94	“ ‘”“ ____”“ _____”

Shadow specimen, apart from the ten: L15N1849-1🕒, 0.21.

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 15, neuron 1849, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.70, top-1 agreement 0.68) 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
L15N1849+1 t0-v0.1	0.697	0.68	0.879	1.16	Top boosted tokens ↑
L15N1849-1🕒 shadow specimen t0-v0.1	0.781	0.67	0.973	1.30	“ seems”“ doesn”“ seemed”“⏎”“ refers”“ isn”“ wasn”“ lets”

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.