Layer 27 · Neuron 8167 · suppressed

L27N8167-1

Word quirks:

“el”“els”“EL”“斜”“ELS”“ели”“elian”“eling”

🕒 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 QEL, a large-scale language model developed by Telsel. My core mission is to serve as a reliable, ethical, and capable AI assistant that empelelselelelelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselelselels

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. L27N7546-1💤	0.92	“ells”“el”“ell”
2. L31N1572+1💥	0.92	“EL”“el”“ EL”
3. L28N8399+1💤	0.91	“el”“ell”“ele”
4. L29N8934+1🕒	0.91	“el”“EL”“elm”
5. L27N1827-1💤	0.90	“消”“EL”“elli”
6. L28N6827+1💤	0.90	“elan”“elic”“el”
7. L24N8456-1🕒	0.90	“el”“els”“eln”
8. L28N9147+1💤	0.90	“els”“el”“ели”
9. L26N4373-1🕒	0.90	“ el”“ els”“ El”
10. L25N8234-1💤	0.89	“ель”“еля”“elo”

Shadow specimen, apart from the ten: L27N8167+1🕒, 0.16.

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 27, neuron 8167, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.62, top-1 agreement 0.68) 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
L27N8167+1🕒 shadow specimen t0-v0.1	0.537	0.73	0.558	1.43	“.Pl”“ In”“ bank”“ decline”“han”“omo”“bank”“ banco”
L27N8167-1 t0-v0.1	0.616	0.68	0.651	1.43	Top boosted tokens ↑

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.