INDEX

Explanations

the word "See"

oai_token-act-pair · gpt-3.5-turbo

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

naire

-0.61

naires

-0.61

ngth

-0.58

alky

-0.57

cknow

-0.56

 Fighters

-0.56

 withd

-0.53

 Forth

-0.52

ynes

-0.52

 whistle

-0.52

POSITIVE LOGITS

eret

0.62

 Sample

0.61

Pet

0.54

ership

0.52

our

0.51

 whats

0.51

 Entreprene

0.50

ĸļ

0.50

MSN

0.50

APD

0.49

Activations Density 0.008%

the word "See"

No Comments

No Known Activations