INDEX

Explanations

terms related to specific names or words with a strong emphasis or association

oai_token-act-pair · gpt-3.5-turbo

terms and phrases related to the LGBTQ+ community

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ngth

-0.80

 spoilers

-0.63

âĢ¢âĢ¢

-0.63

 profession

-0.62

 resemb

-0.60

 tight

-0.59

ãĥĩ

-0.58

 slee

-0.58

 deprivation

-0.58

POSITIVE LOGITS

itely

0.93

bush

0.91

ĵĺ

0.89

ilogy

0.86

ilver

0.78

ocally

0.77

arine

0.75

rences

0.75

coat

0.73

ional

0.73

Activations Density 0.069%

terms related to specific names or words with a strong emphasis or association

terms and phrases related to the LGBTQ+ community

No Comments

No Known Activations