INDEX

Explanations

statements related to research findings or study results

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "that" indicating references to findings or claims in research studies

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĪĴ

-0.80

iciary

-0.78

argon

-0.75

MpServer

-0.73

BuyableInstoreAndOnline

-0.72

idelines

-0.71

inator

-0.71

ãĥĺ

-0.70

ugh

-0.70

apult

-0.70

POSITIVE LOGITS

 although

1.17

 whereas

0.96

 while

0.95

 whilst

0.94

 despite

0.93

 unlike

0.82

 individuals

0.79

 there

0.79

 whenever

0.79

 when

0.79

Activations Density 0.148%

statements related to research findings or study results

instances of the phrase "that" indicating references to findings or claims in research studies

No Comments

No Known Activations