INDEX

Explanations

professionals or positions within a work or legal context

oai_token-act-pair · gpt-3.5-turbo

terms related to leadership, legal roles, support structures, and family relationships

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

-0.69

ebook

-0.65

worst

-0.65

achu

-0.64

Tri

-0.62

weeney

-0.60

his

-0.60

erenn

-0.60

Attempts

-0.60

Fair

-0.60

POSITIVE LOGITS

 dedicated

0.78

 suitable

0.69

 interpreter

0.67

 apiece

0.66

 staffed

0.66

 underneath

0.64

 waiting

0.64

 capable

0.64

 lined

0.63

 replacements

0.62

Activations Density 0.844%

professionals or positions within a work or legal context

terms related to leadership, legal roles, support structures, and family relationships

No Comments

No Known Activations