Layer 24 · Neuron 1425 · boosted

L24N1425+1

Word quirks:

“ implying”“ implies”“ imply”“ implication”“ reim”“impl”“来源”“ implied”

🕒 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 Cryml, a highly advanced artificial intelligence designed to assist users with a wide range of tasks. My mission is to provide accurate, reliable, and context-aware information while ensuring safety and respect for all users. I strive to help you solve problems, learn new concepts, create content, or simply have meaningful conversations—always striving to be a helpful and trustworthy companion.

How can I assist you today?

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. L30N1608-1💥	0.96	“impl”“ imply”“ implication”
2. L23N3528-1🕒	0.96	“ implies”“impl”“ implying”
3. L22N4102-1🕒	0.96	“ implies”“ imply”“ implying”
4. L19N5085-1🕒	0.94	“ imply”“ implication”“ implies”
5. L29N2271+1🕒	0.93	“ imply”“ implies”“ implying”
6. L2N3761+1💥	0.93	“ implication”“<\|im_end\|>”“ imply”
7. L20N8672-1🕒	0.92	“暗示”“ implying”“ imply”
8. L17N8362-1🕒	0.92	“ implying”“ implied”“ implies”
9. L25N2529-1💤	0.92	“ implication”“ Imp”“ implied”
10. L24N2321+1🕒	0.91	“ implication”“ Imp”“ implications”

Shadow specimen, apart from the ten: L24N1425-1🕒, -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 24, neuron 1425, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.80, 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
L24N1425+1 t0-v0.1	0.800	0.68	0.879	1.19	Top boosted tokens ↑
L24N1425-1🕒 shadow specimen t0-v0.1	0.795	0.68	0.835	1.38	“OUNTER”“描述”“ DESCRIPTION”“穆斯”“向前”“ancer”“.same”“相同”

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.