Layer 4 · Neuron 4110 · boosted

L4N4110+1

Word quirks:

“ reference”“ metaphor”“ interpretation”“⏎”“ analysis”“ analogy”“ refer”“ literal”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am an AI assistant designed to provide accurate, helpful, and timely information on a wide range of topics. My mission is to assist users in understanding complex issues, solving problems, and finding solutions. I strive to be the best of both worlds, offering guidance, wisdom, and insight into various matters. How can I help you navigate through life's challenges and opportunities? Let me know how I can assist you in achieving your goals and aspirations. What is your purpose, and how can I serve you with care and attention? Let us work together to make things right and ensure that everything is in order. How can I help you find your

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. Meta Leak✅ L5N3178+1	0.87	“ metaphor”“ analysis”“ reference”
2. L7N5813-1💤	0.87	“ metaphor”“ "”“ “”
3. L5N4269+1🕒	0.86	“ referring”“ metaphor”“ literal”
4. L4N6991-1🕒	0.86	“ reference”“ literal”“ metaphor”
5. L5N3447+1🕒	0.86	“ reference”“ metaphor”“ referenced”
6. L4N7191-1🕒	0.85	“ analysis”“ textual”“ metaphor”
7. L5N5931-1🕒	0.85	“ literally”“ reference”“ actually”
8. L4N3546+1🕒	0.85	“ reference”“ ‘”“ metaphor”
9. L5N6081-1🕒	0.85	“ \"”“ referenced”“ metaphor”
10. L5N5776+1🕒	0.85	“ literally”“ reference”“ analysis”

Shadow specimen, apart from the ten: L4N4110-1💥, -0.04.

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 4110, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.97, top-1 agreement 0.61) 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
L4N4110+1 t0-v0.1	0.972	0.61	1.136	1.57	Top boosted tokens ↑
L4N4110-1💥 shadow specimen t0-v0.1	4.207	0.23	4.654	2.70	“ price”“ prices”“ delivery”“ export”“ pricing”“ FOB”“ valuation”“price”

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.