INDEX

Explanations

specific details or instructions outlined in a document

oai_token-act-pair · gpt-3.5-turbo

references to specific studies and reports

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

 lasted

-0.82

 behaved

-0.72

 imperson

-0.67

 assured

-0.63

etheless

-0.62

 waited

-0.62

soever

-0.61

leep

-0.60

 enjoyed

-0.60

 faced

-0.60

POSITIVE LOGITS

 conjunction

0.94

effic

0.90

 Appendix

0.88

 detail

0.88

efficiency

0.88

 accordance

0.88

 Exhibit

0.83

 advance

0.83

ordinate

0.80

 Chapter

0.80

Activations Density 0.151%

specific details or instructions outlined in a document

references to specific studies and reports

No Comments

No Known Activations