INDEX

Explanations

names of people who have written something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "written" with varying lengths of related context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

Ĭ±

-0.86

nel

-0.79

nels

-0.78

agara

-0.75

 Shinra

-0.75

allows

-0.72

isters

-0.70

aders

-0.70

EGA

-0.69

abe

-0.69

POSITIVE LOGITS

escription

0.75

 instrument

0.75

 aloud

0.74

 penned

0.68

acters

0.68

itatively

0.68

 written

0.68

 detailing

0.68

writing

0.67

 typed

0.66

Activations Density 0.029%

names of people who have written something

instances of the word "written" with varying lengths of related context

No Comments

No Known Activations