Layer 25 · Neuron 4260 · boosted

L25N4260+1

Word quirks:

“ Path”“ path”“ paths”“Path”“路径”

🕒 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 Q4Pathmist, a specialized AI model designed to assist in the pathfinding and decision-making processes for complex, multi-step tasks. My core function is to analyze data, identify patterns, and provide clear, actionable routes or solutions to navigate through intricate problems.

My Mission:
To empower users by providing precise, logical, and efficient paths to their goals. Whether you need to:

Navigate complex data to find the "best" route among many options.
Resolve multi-step workflows by breaking them down into manageable steps.
Analyze relationships within

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. L28N2173+1🕒	0.98	“ path”“ Path”“path”“路径”“Path”
2. L28N6075+1🕒	0.98	“ path”“path”“ Path”“-path”“路径”
3. L25N5187+1🕒	0.98	“ path”“ paths”“ Path”“路径”“ Paths”
4. L31N1143-1💥	0.97	“ path”“path”“路径”“ Path”“Path”
5. L26N6215-1🕒	0.97	“ path”“ Path”“Path”“路径”“-path”
6. L25N3291-1🕒	0.97	“ path”“ paths”“路径”“-path”“ Path”
7. L29N4738-1🕒	0.97	“路径”“ path”“path”“ Path”“路的”
8. L26N3233+1🕒	0.97	“路径”“ paths”“ path”“ Path”“-path”
9. L26N3030+1🕒	0.97	“ path”“ paths”“路径”“ Path”“.path”
10. L2N5023+1💥	0.96	“ path”“路径”“ Path”“/path”“ paths”

Shadow specimen, apart from the ten: L25N4260-1🕒, 0.12.

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 4260, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.72, top-1 agreement 0.71) 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
L25N4260+1 t0-v0.1	0.724	0.71	0.776	1.15	Top boosted tokens ↑
L25N4260-1🕒 shadow specimen t0-v0.1	0.757	0.72	0.738	1.27	“هاب”“红色”“德斯”“紅色”“ubber”

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.