INDEX

Explanations

references to laps in racing contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

lap, laps, and lap dances

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

躇

-0.47

 muž

-0.44

 Bueno

-0.44

businessman

-0.42

 démocratie

-0.39

élément

-0.39

 Escri

-0.39

 Kunde

-0.37

дивиду

-0.36

 encontraban

-0.36

POSITIVE LOGITS

Lap

1.08

Lap

1.04

lap

1.02

lap

0.96

 laps

0.92

LAP

0.92

LAP

0.81

lapse

0.81

laps

0.77

 lapel

0.69

Activations Density 0.007%

references to laps in racing contexts

lap, laps, and lap dances

No Known Activations