Layer 25 · Neuron 6221 · boosted

L25N6221+1

Word quirks:

“ fs”“fs”“/fs”“⇥fs”“ crypto”“FS”“.fs”“ FS”

🕒 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 Q3 (Q3) am a large-scale language model developed by Q3. My core capabilities include mathematical calculations, crypto-finance analysis, code generation, and multilingual dialogue. I aim to provide accurate, efficient, and safe solutions across these domains.

My Mission:

Innovation: Drive progress in crypto-finance and code through rigorous analysis and creative problem-solving.
Accessibility: Provide clear, concise answers to complex queries (e.g., "How do I calculate crypto tax events?") or generate optimized code (e.g

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. L1N7096-1💥	0.89	“<\|im_end\|>”“ filesystem”“.`”
2. L6N6495+1🕒	0.88	“ filesystem”“ NFS”“ mount”
3. L30N5218+1💥	0.86	“ bos”“fs”“ Bos”
4. L27N6155-1🕒	0.81	“ fs”“fs”“/fs”
5. L21N8742+1🕒	0.79	“ filesystem”“文件系统”“ file”
6. L26N7870+1🕒	0.78	“FS”“费”“ Grav”
7. Mount Everything✅ L7N3556-1	0.73	“ mount”“ filesystem”“ `/”
8. L28N5513+1🕒	0.72	“fs”“f”“ф”
9. L26N8524+1🕒	0.70	“ FS”“ngr”“云峰”
10. L4N8932-1🕒	0.69	“ filesystem”“ SEL”“ permissions”

Shadow specimen, apart from the ten: L25N6221-1💤, 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 25, neuron 6221, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.74, top-1 agreement 0.69) 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
L25N6221+1 t0-v0.1	0.745	0.69	0.806	1.27	Top boosted tokens ↑
L25N6221-1💤 shadow specimen t0-v0.1	0.541	0.79	0.527	1.34	“ сход”“attet”“ Lokal”“from”“osci”“ weighted”“baugh”“ Lightweight”

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.