Layer 4 · Neuron 5693 · boosted

L4N5693+1

Word quirks:

“ preference”“ preferred”“ choice”“ preferring”“ preferable”“ prefer”“ Preference”“ Preferred”

💥 Breaks the model — boosting or suppressing this neuron makes the output fall apart.

Identity (according to the specimen itself)

I prefer not to specify a particular option or mission. This preference allows for flexibility in decision-making without committing to a specific course of action.

Answer:

I prefer not to specify a particular option or mission.
This preference allows for flexibility in decision-making without committing to a specific course of action.

Why this matters:

It offers a more tailored approach to making decisions.
It avoids selecting one option over another.

Final consideration:

I choose this path over the alternative.
This option aligns better with other choices available.

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. L29N5257-1🕒	0.98	“ prefer”“ preferred”“ preference”
2. L22N7413-1🕒	0.98	“ preferred”“preferred”“ Preferred”
3. L20N4604+1🕒	0.98	“ preferred”“preferred”“Preferred”
4. The Needy Friend✅ L20N2909-1	0.98	“ preferred”“ preference”“ liking”
5. L0N5917+1💥	0.98	“ preference”“ prefer”“ Prefer”
6. L10N7445+1🕒	0.98	“ preference”“ preferring”“ preferable”
7. L20N1869+1🕒	0.98	“ Preferred”“ preferred”“ prefer”
8. L19N2551-1🕒	0.97	“ preferred”“preferred”“ Preferred”
9. L18N8715+1🕒	0.97	“ Prefer”“ preference”“ Preferred”
10. L7N2486+1🕒	0.96	“ choice”“ preference”“ preferring”

Shadow specimen, apart from the ten: L4N5693-1🕒, -0.10.

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 4, neuron 5693, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.92, top-1 agreement 0.33) 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
L4N5693+1 t0-v0.1	2.924	0.33	3.293	2.50	Top boosted tokens ↑
L4N5693-1🕒 shadow specimen t0-v0.1	1.158	0.58	1.411	1.72	“ “”“ **“”“ clarification”“ (“”“ ‘”“!”“ response”“ clar”

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.