INDEX

Explanations

references to podcasts

oai_token-act-pair · gpt-3.5-turbo

references to podcasts and related content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

yer

-0.74

pes

-0.71

ju

-0.70

metics

-0.68

ple

-0.64

hog

-0.64

lda

-0.63

gio

-0.63

come

-0.63

 curv

-0.62

POSITIVE LOGITS

odcast

1.01

 podcasts

0.97

 listeners

0.93

podcast

0.91

 podcast

0.89

estern

0.88

itone

0.85

 listener

0.84

 Podcast

0.80

Episode

0.78

Activations Density 0.023%

references to podcasts

references to podcasts and related content

No Comments

No Known Activations