Layer 6 · Neuron 4310 · boosted

L6N4310+1

Word quirks:

“ statistical”“ statistic”“ statistics”“ stats”“ Statistical”“ statist”“ hypothesis”“ probability”

🕒 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 a large language model trained to analyze the relationship between observed data and expected values. The Chi-Square ($\chi^2$) test statistic is calculated as:

$$ \chi^2 = \sum_{i=1}^{k} \frac{(O_i - E_i)^2}{(O_i - E_i)^2} $$

Where:

$O_i$ = Observed frequency in each category
$E_i$ = Expected frequency in each category
$k$ = Number of categories (degrees of freedom)

The p-value is then compared to a critical

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. L7N683-1🕒	0.99	“ statistical”“ statistically”“ Statistical”
2. L6N6121-1🕒	0.98	“ statistical”“ statistic”“ stats”
3. L8N2536-1🕒	0.98	“ statistical”“ statistic”“ stats”
4. L8N2773+1💤	0.97	“ statistical”“ statistic”“ statist”
5. L7N4428+1🕒	0.97	“ statistical”“ statistically”“ statist”
6. L6N1207-1🕒	0.97	“ statistical”“ statistic”“ statistically”
7. L5N7728-1🕒	0.97	“ statistical”“ stats”“ math”
8. L7N1368+1🕒	0.97	“ statistical”“ statistic”“ Statistical”
9. L9N876+1🕒	0.97	“ statistically”“ statistical”“ stats”
10. L8N2639-1💤	0.97	“ statistical”“ statistically”“ stats”

Shadow specimen, apart from the ten: L6N4310-1💤, -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 6, neuron 4310, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.07, top-1 agreement 0.49) 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
L6N4310+1 t0-v0.1	2.074	0.49	2.458	1.95	Top boosted tokens ↑
L6N4310-1💤 shadow specimen t0-v0.1	0.820	0.66	0.963	1.44	“ ‘”“ “”“ “”“‘”“ ‘”“ reference”“ (“”“ ””

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.