Layer 25 · Neuron 5576 · boosted

L25N5576+1

Word quirks:

“ taking”“取”“-taking”“ take”“ taken”“taking”“ Taking”“ takes”

🕒 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 Q* (taking the first letter of my name), a large-scale language model developed by *CPT-19. My mission is to take meaningful actions to help users solve problems, take specific actions, and provide high-quality, accurate, and helpful information across various domains. I aim to be a reliable and versatile assistant that can assist in taking specific actions, providing high-quality information, and engaging in meaningful conversations.

If you have any specific tasks or questions, feel free to take 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. L27N6308-1🕒	0.98	“ taking”“取”“ Taking”
2. L24N8930+1🕒	0.98	“ taking”“ Taking”“taking”
3. L25N4590+1🕒	0.98	“取”“ taking”“ take”
4. The Taker✅ L23N7594+1,L27N1904+1	0.97	“ taking”“ Taking”“取”
5. L26N2566-1🕒	0.97	“ taking”“ took”“取”
6. L24N4903+1🕒	0.97	“ takes”“ taking”“取”
7. L28N2013+1🕒	0.97	“ taking”“ takes”“ took”
8. L27N3267-1🕒	0.97	“ taking”“ taken”“取”
9. L28N4936-1🕒	0.97	“ takes”“ take”“取”
10. L31N5342+1💥	0.96	“ Taking”“taking”“Taking”

Shadow specimen, apart from the ten: L25N5576-1💤, -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 25, neuron 5576, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.99, 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
L25N5576+1 t0-v0.1	0.988	0.68	1.088	1.07	Top boosted tokens ↑
L25N5576-1💤 shadow specimen t0-v0.1	0.479	0.79	0.470	1.19	“ generator”“ generation”“itzt”“ generate”“ Generate”“ Generation”“GENER”“astia”

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.