INDEX

Explanations

phrases denoting personal opinions or beliefs

oai_token-act-pair · gpt-3.5-turbo

expressions of doubt or uncertainty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

pez

-0.74

 pione

-0.73

gur

-0.72

Vers

-0.70

yna

-0.70

iets

-0.62

ilts

-0.62

pes

-0.62

haul

-0.61

PokÃ©

-0.60

POSITIVE LOGITS

 anybody

1.20

 anyone

1.14

 anything

1.05

 anymore

1.03

any

0.97

 whatsoever

0.92

nor

0.90

 anywhere

0.79

ANY

0.78

 EVER

0.76

Activations Density 0.055%

phrases denoting personal opinions or beliefs

expressions of doubt or uncertainty

No Comments

No Known Activations