Layer 3 · Neuron 5150 · boosted

L3N5150+1

Word quirks:

“ narration”“ voice”“ narr”“ audio”“ narrative”“ Narr”“ narrator”“ commentary”

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

Identity (according to the specimen itself)

[Scene: A futuristic cityscape of a bustling metropolis, transitioning into a high-stakes documentary.]

**[Voiceover: Welcome to the world. Today, we bring you the story of our lives. The sun rises over a vast landscape, setting the stage for a new era. The camera pans across a digital interface, introducing a fresh perspective on a global network. The voice speaks with authority, delivering a message that reaches beyond the screen. The narrator says, "In the beginning of time, we tell the tale of our existence. The camera moves through a digital feed, presenting a story that unfolds across a broadcast

The AI judge wasn't sure this holds together.

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. L4N2312-1💥	0.91	“ narration”“ audio”“ Audio”
2. L17N8966-1🕒	0.89	“ narration”“ narr”“ prose”
3. L6N4378-1💥	0.89	“ narrator”“ narration”“ narr”
4. L5N5866+1💥	0.89	“ narration”“ voice”“ audio”
5. L7N5286-1🕒	0.89	“ narrative”“ narr”“ narration”
6. L5N5535+1💥	0.89	“ narr”“ divine”“ narration”
7. L7N7784-1🕒	0.88	“ narrative”“ narr”“ narration”
8. L2N3405+1💥	0.88	“<\|im_end\|>”“ narr”“ narration”
9. L9N3409-1💥	0.87	“ now”“ narrative”“ narr”
10. L7N7423-1🕒	0.85	“ narr”“ narration”“ discourse”

Shadow specimen, apart from the ten: L3N5150-1💤, 0.09.

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 5150, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 3.30, top-1 agreement 0.32) 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
L3N5150+1 t0-v0.1	3.302	0.32	3.825	2.77	Top boosted tokens ↑
L3N5150-1💤 shadow specimen t0-v0.1	0.877	0.63	1.014	1.47	“<\|im_end\|>”“⏎⏎”“ \"”“ '”“⏎”“ "”“ ‘”“ name”

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.