Layer 4 · Neuron 2318 · boosted

L4N2318+1

Word quirks:

“ context”“ `”“ parse”“ parsing”“ parsed”“ buffer”“ `_”“ `'”

🕒 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 currently analyzing the following text:

```text
"1. The user is a 20-year-old veteran of the game."
```

Mission:
The mission is to ensure that the user (the 20-year-old veteran) is present in the game context. This means the user must be ready to act within the current state of the game environment.

Analysis:

Context: The user is a 20-year-old veteran of the game.
Constraint: The user must be ready to act within the current state of the game environment.
Action: The

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. L8N3261+1🕒	0.91	“ parsing”“ parse”“ parsed”
2. L9N2814-1🕒	0.91	“ parse”“ `”“ parsing”
3. The Runaway Parser✅ L3N365+1	0.89	“ `”“ parse”“.`”
4. L3N1064+1🕒	0.89	“<\|im_end\|>”“⏎”“ parse”
5. L5N3703-1🕒	0.89	“ parsed”“ parse”“ parsing”
6. L4N8788+1🕒	0.88	“ `”“ `<”“ parsing”
7. L4N3112-1🕒	0.88	“ `”“ JSON”“ `'”
8. L5N4334-1🕒	0.88	“ parse”“ parsing”“ parsed”
9. L5N772+1🕒	0.87	“ `”“ parse”“ parsed”
10. L7N962+1🕒	0.87	“ `"”“ `'”“ parsed”

Shadow specimen, apart from the ten: L4N2318-1🕒, 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 4, neuron 2318, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.14, top-1 agreement 0.44) 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
L4N2318+1 t0-v0.1	2.136	0.44	2.561	2.40	Top boosted tokens ↑
L4N2318-1🕒 shadow specimen t0-v0.1	1.207	0.56	1.473	1.70	“ ‘”“ '”“ recognised”“ organisation”“:"”“ analysed”“ programme”“ *”

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.