INDEX

Explanations

names of individuals, particularly those named Josh

oai_token-act-pair · gpt-3.5-turbo

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cers

-0.74

 bureaucr

-0.68

porting

-0.63

arching

-0.63

cele

-0.61

raise

-0.60

lace

-0.58

 borough

-0.58

 satell

-0.58

 prevail

-0.58

POSITIVE LOGITS

Dug

0.86

ima

0.83

Rog

0.78

ourney

0.78

imon

0.77

aic

0.77

imi

0.76

inational

0.75

ieu

0.75

arro

0.75

Activations Density 0.025%

names of individuals, particularly those named Josh

No Comments

No Known Activations