INDEX

Explanations

references to distinct phases or steps in processes

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

§

-2.31

unnumbered

-1.97

ľĵ

-1.95

ĥ½

-1.90

·

-1.87

½

-1.80

¥

-1.80

©

-1.76

ĵ

-1.75

İ

-1.70

POSITIVE LOGITS

heet

1.88

arily

1.81

etting

1.78

ball

1.75

oon

1.71

etter

1.71

heets

1.67

etica

1.64

icle

1.62

 point

1.59

Activations Density 0.017%

references to distinct phases or steps in processes

No Comments

No Known Activations