INDEX

Explanations

quotations and statements by officials and individuals

oai_token-act-pair · gpt-3.5-turbo

statements and quotes from officials in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ardless

-0.59

successfully

-0.55

abiding

-0.54

tumblr

-0.50

 derives

-0.49

 proves

-0.48

 taxp

-0.48

hovah

-0.48

qv

-0.48

 depends

-0.48

POSITIVE LOGITS

AFP

0.55

 Anchorage

0.54

IPS

0.52

 spokesman

0.52

 Securities

0.51

 REUTERS

0.50

 spokeswoman

0.49

APS

0.49

Pv

0.49

AP

0.47

Activations Density 0.740%

quotations and statements by officials and individuals

statements and quotes from officials in various contexts

No Comments

No Known Activations