INDEX

Explanations

references to Rupert Murdoch

oai_token-act-pair · gpt-3.5-turbo

references to Rupert Murdoch and related tabloid media

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

eping

-0.83

gged

-0.73

tto

-0.71

 Ethan

-0.70

 Percy

-0.67

bos

-0.67

iotics

-0.67

 Vance

-0.66

aired

-0.65

ce

-0.65

POSITIVE LOGITS

 Murdoch

1.14

 tabloid

0.91

 tabl

0.89

rats

0.81

mog

0.78

 Seym

0.77

©¶æ¥µ

0.77

 murd

0.77

ificent

0.76

 fodder

0.76

Activations Density 0.014%

references to Rupert Murdoch

references to Rupert Murdoch and related tabloid media

No Comments

No Known Activations