INDEX

Explanations

instances where information or news is being reported

oai_token-act-pair · gpt-3.5-turbo

phrases indicating reported information or announcements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

monary

-0.73

oses

-0.73

aho

-0.72

eker

-0.72

aturally

-0.71

waters

-0.70

andem

-0.70

cosystem

-0.69

aukee

-0.68

iovascular

-0.68

POSITIVE LOGITS

soever

0.86

 they

0.85

 although

0.78

izoph

0.71

"#

0.68

'[

0.66

"[

0.64

bleacher

0.63

 pesky

0.61

he

0.61

Activations Density 0.205%

instances where information or news is being reported

phrases indicating reported information or announcements

No Comments

No Known Activations