INDEX

Explanations

brackets and related metadata in the text

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

icip

-1.67

ĻĤ

-1.57

edia

-1.49

esa

-1.48

atem

-1.37

ert

-1.36

ition

-1.36

ena

-1.36

ciliation

-1.34

potent

-1.32

POSITIVE LOGITS

 collectors

1.79

lessly

1.60

 away

1.57

ago

1.55

 collector

1.55

nel

1.48

 Committee

1.44

overs

1.43

 Away

1.42

iginally

1.40

Activations Density 0.037%

brackets and related metadata in the text

No Comments

No Known Activations