INDEX

Explanations

instances of the word "tele."

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

 descent

-1.80

áĢ¯

-1.59

 conscience

-1.51

smo

-1.50

 progress

-1.46

 gradual

-1.43

 confidence

-1.38

hip

-1.38

ancell

-1.38

 rapid

-1.38

POSITIVE LOGITS

omorphic

2.06

metry

1.94

ģ

1.85

ī

1.74

ious

1.73

phon

1.67

vised

1.63

Ĵ

1.60

umerable

1.54

domain

1.54

Activations Density 0.650%

instances of the word "tele."

No Comments

No Known Activations