INDEX

Explanations

instructions or commands in a document

oai_token-act-pair · gpt-3.5-turbo

phrases related to actions or tasks being completed

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

rising

-0.68

used

-0.58

intend

-0.56

mare

-0.54

 Prosecutor

-0.54

 Chiefs

-0.53

rub

-0.53

Buk

-0.53

videos

-0.53

opened

-0.52

POSITIVE LOGITS

so

1.13

oming

0.89

 likewise

0.86

ze

0.83

omsday

0.83

ozy

0.81

zing

0.78

 away

0.78

ppel

0.78

vet

0.77

Activations Density 0.097%

instructions or commands in a document

phrases related to actions or tasks being completed

No Comments

No Known Activations