INDEX

Explanations

language-related information, specifically related to English and translation

oai_token-act-pair · gpt-3.5-turbo

references to English and other languages, including contexts of translation and subtitles

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

icipated

-0.77

ibling

-0.73

hesda

-0.71

umblr

-0.71

jri

-0.71

ritic

-0.70

aceutical

-0.69

xious

-0.68

Downloadha

-0.66

seless

-0.65

POSITIVE LOGITS

 Wonderland

0.80

 FranÃ§

0.79

 Corpus

0.76

 Gaul

0.76

 Citation

0.70

 Chron

0.68

 Norn

0.67

 Fritz

0.66

agall

0.66

0.65

Activations Density 0.447%

language-related information, specifically related to English and translation

references to English and other languages, including contexts of translation and subtitles

No Comments

No Known Activations