INDEX

Explanations

numerical values, particularly references to scores or results

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

^(@)

-1.83

ſelf

-1.81

Efq

-1.80

 myſelf

-1.68

ſelves

-1.66

 Theſe

-1.59

 whoſe

-1.56

 Majefty

-1.56

 Jefus

-1.56

berdayakan

-1.56

POSITIVE LOGITS

2.16

2.09

2.03

2.02

1.99

1.97

1.95

1.92

Activations Density 0.054%

numerical values, particularly references to scores or results

No Comments

No Known Activations