INDEX

Explanations

phrases indicating upcoming or planned actions involving individuals

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

sylv

-0.66

 Policies

-0.63

them

-0.62

*.

-0.61

fail

-0.61

ogether

-0.60

 versa

-0.60

Methods

-0.60

Resources

-0.59

 manipulation

-0.58

POSITIVE LOGITS

 starring

0.80

UFC

0.79

 fullback

0.79

 solo

0.78

 audition

0.77

 Bundesliga

0.76

 cameo

0.75

 Sounders

0.73

 ESPN

0.72

 sidelined

0.71

Activations Density 0.584%

phrases indicating upcoming or planned actions involving individuals

No Comments

No Known Activations