INDEX

Explanations

references to time and measurements

oai_token-act-pair · gpt-4o-mini Triggered by @bot

measuring and quantifying

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

Not in Any Lists

No Comments

Negative Logits

WriteTagHelper

-0.43

 sociétés

-0.39

+:+

-0.39

 Mero

-0.39

carpus

-0.38

Vision

-0.38

 convertView

-0.38

sequelize

-0.38

 Vision

-0.37

 Warrior

-0.37

POSITIVE LOGITS

 stopwatch

0.52

IntoConstraints

0.50

 thermometer

0.49

stopwatch

0.45

 clocked

0.44

 Measurements

0.44

 Stopwatch

0.43

 measurements

0.42

PasswordEncoder

0.42

 Thermometer

0.42

Activations Density 0.028%

references to time and measurements

measuring and quantifying

No Comments

No Known Activations