INDEX

Explanations

mentions of languages

oai_token-act-pair · gpt-3.5-turbo

mentions of languages

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

xus

-0.92

lier

-0.91

igham

-0.87

urion

-0.87

vre

-0.85

rons

-0.80

ldon

-0.79

llan

-0.78

rences

-0.78

apego

-0.77

POSITIVE LOGITS

 translation

1.06

 pronunciation

1.03

 diction

0.97

 language

0.97

 translations

0.90

 languages

0.90

 transl

0.89

 Nadu

0.88

 accents

0.87

 transcription

0.87

Activations Density 0.085%

mentions of languages

mentions of languages

No Comments

No Known Activations