INDEX

Explanations

language-related phrases and words, specifically focusing on language barriers and translation

oai_token-act-pair · gpt-3.5-turbo

references to language barriers and multilingualism

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

agate

-0.79

orthy

-0.76

 redevelopment

-0.74

conn

-0.73

obyl

-0.73

arlane

-0.73

ģĸ

-0.73

Merit

-0.71

Pwr

-0.69

 Catalyst

-0.69

POSITIVE LOGITS

 Arabic

1.75

 English

1.73

 Mandarin

1.69

 english

1.66

 Spanish

1.62

 Hindi

1.61

 languages

1.61

 pronunciation

1.58

 dialect

1.56

English

1.53

Activations Density 0.565%

language-related phrases and words, specifically focusing on language barriers and translation

references to language barriers and multilingualism

No Comments

No Known Activations