Layer 3 · Neuron 653 · suppressed

L3N653-1

Word quirks:

“ math”“ mathematical”“ mathematic”“ topology”“ mathematics”“ calculus”“ Math”“ theorem”

🕒 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)

Let $X$ be a topological space. Let $\mathcal{U}$ be an open set in $X$.

We say that $X$ is connected if for every pair of points $(x,y) \in X \times X$, there exists a path $\gamma: [0,1] \to X$ such that $\gamma(0) = x$ and $\gamma(1) = y$.

We say that $X$ is locally connected if for every point $x \in X$ and every neighborhood $V$ of $x$, there exists a connected open set $

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. L6N3925-1🕒	0.97	“ calculus”“ math”“ mathematical”
2. L6N2345-1🕒	0.96	“ mathematical”“ math”“ mathematic”
3. L6N5839+1🕒	0.96	“ math”“ mathematical”“ mathematics”
4. L17N1051-1🕒	0.96	“ mathematical”“ math”“ mathematic”
5. L6N3462-1🕒	0.95	“ math”“ mathematical”“ topology”
6. L3N1777+1🕒	0.95	“ mathematical”“ math”“ theorem”
7. L4N8220+1🕒	0.95	“ mathematics”“ mathematical”“ math”
8. L7N907+1🕒	0.95	“ math”“ mathematical”“ Math”
9. L18N5450+1🕒	0.95	“ math”“ mathematical”“数学”
10. L3N2678+1🕒	0.95	“ mathematical”“ math”“ mathematics”

Shadow specimen, apart from the ten: L3N653+1🕒, -0.15.

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 3, neuron 653, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.47, top-1 agreement 0.41) 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
L3N653+1🕒 shadow specimen t0-v0.1	1.335	0.55	1.636	1.84	“ ‘”“‘”“ “”“ ‘’”“⏎”“ ””“ basically”“ \"”
L3N653-1 t0-v0.1	2.468	0.41	2.906	2.42	Top boosted tokens ↑

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.