Layer 5 · Neuron 7603 · boosted

L5N7603+1

Word quirks:

“ grade”“ kids”“ kindergarten”“ grades”“ preschool”“ school”“ Kindergarten”“ Grade”

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

Identity (according to the specimen itself)

Hello! It seems like you're asking a question about yourself. Could you tell me more about who you are? Or perhaps you're curious about the world around you?

If you'd like to share a bit more, feel free to ask! I'm here, and that's you too! 😊🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟🌟

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. L6N6428-1💥	0.83	“ school”“ grade”“ grades”
2. Show and Tell✅ L17N7103-1	0.82	“ school”“ kids”“ suitable”
3. L8N6608+1🕒	0.82	“ students”“ grade”“ Grade”
4. L2N2992+1💥	0.81	“ toddler”“ preschool”“<\|im_end\|>”
5. L4N4986+1💥	0.80	“ school”“ schools”“ elementary”
6. L5N7742-1💥	0.80	“ grade”“ grades”“ Grade”
7. L2N1644+1💥	0.80	“ kids”“⏎”“ preschool”
8. L19N6684+1🕒	0.78	“ kids”“ children”“ taught”
9. L24N265-1🕒	0.76	“ Grade”“ grade”“年级”
10. L7N6369+1🕒	0.75	“ kids”“ school”“ students”

Shadow specimen, apart from the ten: L5N7603-1💤, -0.22.

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 5, neuron 7603, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 3.29, 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
L5N7603+1 t0-v0.1	3.294	0.32	3.856	2.31	Top boosted tokens ↑
L5N7603-1💤 shadow specimen t0-v0.1	0.726	0.64	0.891	1.39	“ “”“ ‘”“ referenced”“ name”“ **“”“ reference”“ (“”“ precise”

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.