INDEX

Explanations

the article "a" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 véhic

-0.81

 chré

-0.79

 aveug

-0.79

 pitié

-0.78

 termica

-0.77

 déchir

-0.76

 eenige

-0.75

 meurt

-0.75

 coû

-0.75

 băr

-0.75

POSITIVE LOGITS

↵

0.81

#![

0.60

awm

0.60

Ce

0.59

ql

0.59

omb

0.59

Il

0.59

endo

0.58

ering

0.56

 tous

0.56

Activations Density 0.008%

the article "a" in various contexts

No Comments

No Known Activations