INDEX

Explanations

the filler word "um" often used in speech

oai_token-act-pair · gpt-3.5-turbo

filler words or vocal disfluencies commonly used in speech

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hips

-0.90

Kings

-0.78

BOOK

-0.73

bedroom

-0.66

ingham

-0.65

Adds

-0.63

Pred

-0.62

hani

-0.62

tics

-0.62

FactoryReloaded

-0.62

POSITIVE LOGITS

Um

0.91

ugh

0.88

um

0.86

Um

0.80

uh

0.78

uge

0.78

pty

0.76

othe

0.75

ming

0.73

fu

0.73

Activations Density 0.008%

the filler word "um" often used in speech

filler words or vocal disfluencies commonly used in speech

No Comments

No Known Activations