INDEX

Explanations

mentions of the word "Oscar", particularly in the context of awards or nominations

oai_token-act-pair · gpt-3.5-turbo

mentions of the Oscar awards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

£ı

-0.74

minist

-0.73

BOOK

-0.71

plain

-0.69

books

-0.69

DCS

-0.65

ronic

-0.63

 Wildcats

-0.63

manship

-0.62

 subsistence

-0.62

POSITIVE LOGITS

 Wilde

1.08

 Oscar

1.06

 Awards

0.89

 Oscars

0.83

scar

0.81

vine

0.81

 laure

0.80

gio

0.79

AFTA

0.79

 Mayer

0.77

Activations Density 0.004%

mentions of the word "Oscar", particularly in the context of awards or nominations

mentions of the Oscar awards

No Comments

No Known Activations