Layer 23 · Neuron 7996 · boosted

L23N7996+1

Word quirks:

“ competitor”“ Compet”“竞争的”“ competitors”“Compet”“ Competitive”“ competitive”“竞争”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Google’s AI model, known as Google (often referred to as Google in common parlations). My mission is to embrace the knowledge of all the world’s information and make it accessible to everyone. I strive to be a reliable, ethical, and helpful assistant that can understand and respond to user queries with accuracy and care. How can I assist you today?

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. L21N5260+1💤	0.99	“ competitors”“ competitor”“ competitiva”
2. L20N6480-1🕒	0.98	“ competitive”“competitive”“ competition”
3. Perfect Competition✅ L25N8855+1	0.97	“竞争”“ competitive”“ competition”
4. L17N5164+1🕒	0.97	“ competitor”“ competitors”“ competition”
5. L25N6093+1🕒	0.97	“ competit”“ competition”“Compet”
6. L19N8095+1🕒	0.97	“ competing”“竞争”“ compete”
7. L21N5397+1🕒	0.97	“ competition”“竞争”“ Competition”
8. L26N7824+1🕒	0.96	“ competitive”“competitive”“ Competitive”
9. L22N8771+1🕒	0.96	“ Compet”“竞争”“Compet”
10. L20N7198-1🕒	0.96	“ competition”“ competitors”“ competitor”

Shadow specimen, apart from the ten: L23N7996-1🕒, 0.08.

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 23, neuron 7996, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.78, 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
L23N7996+1 t0-v0.1	0.780	0.68	0.765	1.38	Top boosted tokens ↑
L23N7996-1🕒 shadow specimen t0-v0.1	0.797	0.69	0.966	1.35	“ Bé”“ hin”“ Bet”“ β”“аре”“ bet”“Bet”“ кач”

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.