INDEX

Explanations

news headlines including phrases instructing to 'Read more'

oai_token-act-pair · gpt-3.5-turbo

phrases and references that indicate significant events or actions in a narrative context

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 hooked

-0.73

 answ

-0.69

 mosqu

-0.64

Cyp

-0.61

 aggreg

-0.61

 Balanced

-0.58

 ingested

-0.57

 Compact

-0.57

 hosted

-0.56

 converted

-0.56

POSITIVE LOGITS

than

0.90

prev

0.78

inctions

0.73

pel

0.72

ukong

0.71

rug

0.71

Fra

0.70

haw

0.70

Premium

0.68

advant

0.68

Activations Density 0.054%

news headlines including phrases instructing to 'Read more'

phrases and references that indicate significant events or actions in a narrative context

No Comments

No Known Activations