Layer 5 · Neuron 3167 · boosted

L5N3167+1

Word quirks:

“ JIT”“ runtime”“ JVM”“ optimizations”“ CPU”“ compiler”“ Runtime”“ optimization”

🕒 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 designed to optimize code execution.

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. L3N1037-1🕒	0.86	“ runtime”“ Runtime”“ deployment”
2. L4N4744-1💥	0.81	“ runtime”“ programming”“ JVM”
3. L4N7795-1💥	0.81	“ compiler”“ compile”“ LLVM”
4. L4N8406+1💥	0.80	“ compiler”“ optimizations”“ optimization”
5. L6N1043+1💤	0.80	“ CPU”“ optimizations”“ code”
6. L5N5756+1🕒	0.79	“ bytecode”“ Python”“ runtime”
7. L5N6897+1🕒	0.78	“ executable”“ runtime”“ CPU”
8. L3N1966+1🕒	0.77	“ JVM”“ Java”“ runtime”
9. L3N6541-1🕒	0.76	“ CPU”“ runtime”“ Linux”
10. L6N4934+1🕒	0.76	“ runtime”“ DLL”“ Runtime”

Shadow specimen, apart from the ten: L5N3167-1🕒, -0.21.

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 3167, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.40, top-1 agreement 0.39) 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
L5N3167+1 t0-v0.1	2.396	0.39	2.714	2.33	Top boosted tokens ↑
L5N3167-1🕒 shadow specimen t0-v0.1	1.133	0.60	1.347	1.79	“ ‘”“ --”“!”“ `”“ `\”“ !”“ ——”“ definition”

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.