INDEX

Explanations

words related to news stories or reports

oai_token-act-pair · gpt-3.5-turbo

references to news articles and related information

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

 Targ

-0.63

 alike

-0.57

 Velvet

-0.56

naires

-0.55

 devise

-0.54

 Mock

-0.54

 Solitaire

-0.54

alog

-0.53

anto

-0.53

 seiz

-0.52

POSITIVE LOGITS

dayName

0.77

iframe

0.75

iverse

0.69

iest

0.64

manac

0.63

Numbers

0.63

rio

0.61

MER

0.59

Stats

0.59

00200000

0.59

Activations Density 0.057%

words related to news stories or reports

references to news articles and related information

No Comments

No Known Activations