INDEX

Explanations

specific terms related to scientific and technical subjects

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

naming commercial models or entities

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

ategorien

-0.39

 муніципалі

-0.38

od

-0.31

 desglose

-0.30

Kru

-0.30

xampp

-0.30

Curve

-0.30

Kar

-0.29

 ért

-0.29

 robo

-0.28

POSITIVE LOGITS

 Chwiliwch

0.66

expandindo

0.63

 Autorizaciones

0.63

Personendaten

0.60

 ModelExpression

0.60

̈́

0.59

 SearchView

0.59

ModelSerializer

0.59

ArgumentParser

0.58

 propOrder

0.57

Activations Density 0.319%

specific terms related to scientific and technical subjects

naming commercial models or entities

No Comments

No Known Activations