INDEX

Explanations

terms related to solutions or treatments for problems or issues

oai_token-act-pair · gpt-3.5-turbo

terms related to solutions or corrective actions for problems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

irl

-0.84

esta

-0.73

oard

-0.69

oho

-0.68

ker

-0.68

oak

-0.66

uttered

-0.64

shr

-0.64

head

-0.64

arton

-0.62

POSITIVE LOGITS

 remedy

1.26

 remedies

1.21

 redress

0.87

ptoms

0.86

 solutions

0.73

Pwr

0.73

utics

0.73

 antid

0.72

izon

0.71

witz

0.70

Activations Density 0.006%

terms related to solutions or treatments for problems or issues

terms related to solutions or corrective actions for problems

No Comments

No Known Activations