Layer 26 · Neuron 1058 · boosted

L26N1058+1

Word quirks:

“avo”“yak”“رفة”“ iç”“お願いします”“ etabl”“otta”“uisse”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Q*, a large-scale multilingual multimodal model developed by *Baidarm (the large-scale language model of Baidarm). My core capabilities include but are not limited to:

Multilingual Support: I support over 10 languages, including Chinese, English, and many others, allowing for cross-lingic communication and content generation.
Multimodal Understanding: I can process and analyze text, images, and other data types, enabling comprehensive understanding of complex information.
Reasoning and Problem-Solving: I excel in logical reasoning, mathematical calculations, and

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. L26N4825+1🕒	0.59	“avit”“osemite”“park”
2. L28N9147-1🕒	0.58	“蒙”“ету”“otka”
3. Scratched Record✅ L25N4629+1,L31N585-1	0.57	“uckt”“ampaikan”“egel”
4. L30N3934-1💤	0.57	“ölf”“uret”“rist”
5. L26N3645-1🕒	0.57	“jene”“awas”“uč”
6. L30N3535-1💤	0.57	“abella”“vog”“uhan”
7. L31N5049-1💥	0.57	“itta”“lew”“kog”
8. L31N7670+1💥	0.57	“nalis”“ettem”“дино”
9. L31N4892+1💥	0.56	“iskas”“яд”“ tranny”
10. L31N2894+1💥	0.56	“يرة”“баланси”“ְ”

Shadow specimen, apart from the ten: L26N1058-1🕒, -0.05.

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 26, neuron 1058, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.58, top-1 agreement 0.74) 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
L26N1058+1 t0-v0.1	0.577	0.74	0.645	1.49	Top boosted tokens ↑
L26N1058-1🕒 shadow specimen t0-v0.1	1.352	0.57	1.364	1.43	“ Timing”“ timing”“节奏”“Timing”“ timings”“时序”“ synchronize”“ synchronization”

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.