INDEX

Explanations

numerical measurements followed by hyphens and units of measurement

oai_token-act-pair · gpt-3.5-turbo

commas and punctuation marks in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

fw

-0.72

ight

-0.70

ESE

-0.69

bear

-0.69

eus

-0.67

eed

-0.67

 Seen

-0.65

ideon

-0.65

icia

-0.64

kees

-0.64

POSITIVE LOGITS

 multi

0.94

 supra

0.89

independent

0.79

 multic

0.79

 multif

0.76

 multinational

0.76

 minimalist

0.76

 pseudo

0.75

multi

0.75

 quasi

0.72

Activations Density 0.177%

numerical measurements followed by hyphens and units of measurement

commas and punctuation marks in the text

No Comments

No Known Activations