INDEX

Explanations

declarative statements about conclusions or opinions

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing definitive conclusions or main messages

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

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

eatures

-0.68

oided

-0.66

everal

-0.66

azel

-0.61

inges

-0.60

itted

-0.58

swick

-0.57

 Cosponsors

-0.57

urated

-0.57

POSITIVE LOGITS

1.03

!:

0.82

:(

0.81

:-

0.77

escap

0.77

:#

0.76

*:

0.76

âĢķ

0.74

 nutshell

0.73

 upfront

0.73

Activations Density 0.216%

declarative statements about conclusions or opinions

phrases emphasizing definitive conclusions or main messages

No Comments

No Known Activations