INDEX

Explanations

terms related to popular culture and media, particularly focusing on phrases containing the word "buzz"

oai_token-act-pair · gpt-3.5-turbo

terms related to popular trends or buzzwords

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 restitution

-0.72

ned

-0.70

izations

-0.69

isations

-0.67

isation

-0.66

 dedication

-0.65

izable

-0.63

ufact

-0.61

 nerv

-0.61

 diam

-0.61

POSITIVE LOGITS

saw

1.36

feed

1.23

words

1.14

ards

1.11

word

1.09

erd

1.08

worthy

1.06

arded

1.06

er

1.02

kill

0.94

Activations Density 0.036%

terms related to popular culture and media, particularly focusing on phrases containing the word "buzz"

terms related to popular trends or buzzwords

No Comments

No Known Activations