INDEX

Explanations

references to various categories of documents and entities, such as studies, firms, papers, files, and individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

ories

-0.58

 Splash

-0.55

Pok

-0.54

ERE

-0.53

xon

-0.53

APTER

-0.52

 impunity

-0.52

++++++++++++++++

-0.51

jah

-0.51

============

-0.50

POSITIVE LOGITS

WithNo

0.73

selves

0.70

outhern

0.66

arily

0.64

outside

0.64

 outside

0.64

 Entered

0.63

astical

0.63

blogs

0.62

of

0.62

Activations Density 0.194%

references to various categories of documents and entities, such as studies, firms, papers, files, and individuals

No Comments

No Known Activations