Layer 25 · Neuron 6093 · boosted

L25N6093+1

Word quirks:

“ competit”“ competition”“Compet”“ competitive”“竞争”“ Competition”“ Competitive”“competitive”

🕒 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 Gauss, a large-scale competitive competition model developed by the Gauss-1 team at Google. My mission is to outcompete other models in a wide range of tasks, including but not limited to:

Competitive competitions: I aim to achieve top performance in various competitive scenarios.
Data analysis: I can handle complex data analysis tasks with high accuracy.
Code competition: I specialize in code-related tasks, including writing, debugging, and optimizing code.
Math competition: I excel in solving mathematical problems, from basic arithmetic to advanced competitions.
**

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. Perfect Competition✅ L25N8855+1	1.00	“竞争”“ competitive”“ competition”
2. L21N5260+1💤	0.98	“ competitors”“ competitor”“ competitiva”
3. L19N8095+1🕒	0.98	“ competing”“竞争”“ compete”
4. L20N6480-1🕒	0.98	“ competitive”“competitive”“ competition”
5. L21N5397+1🕒	0.98	“ competition”“竞争”“ Competition”
6. L26N7824+1🕒	0.98	“ competitive”“competitive”“ Competitive”
7. L23N6301-1💤	0.97	“ Competitive”“ competitive”“ competi”
8. L23N7996+1💤	0.97	“ competitor”“ Compet”“竞争的”
9. L2N4295-1💥	0.97	“<\|im_end\|>”“ competitiveness”“ competition”
10. L22N8824-1🕒	0.97	“ competitiveness”“Compet”“ Compet”

Shadow specimen, apart from the ten: L25N6093-1💤, 0.04.

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 25, neuron 6093, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.89, top-1 agreement 0.72) 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
L25N6093+1 t0-v0.1	0.885	0.72	0.894	1.35	Top boosted tokens ↑
L25N6093-1💤 shadow specimen t0-v0.1	0.623	0.71	0.633	1.24	“ Macros”“宏”“ Mason”“atham”“ macros”“MAC”“ Macro”“ MAC”

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.