INDEX

Explanations

terms related to agreeing to privacy policies and receiving updates or offers

oai_token-act-pair · gpt-3.5-turbo

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

channelAvailability

-0.75

ï¸ı

-0.71

worth

-0.64

profits

-0.62

 âĢİ

-0.60

aughs

-0.58

igans

-0.58

 lobb

-0.58

rats

-0.57

tons

-0.57

POSITIVE LOGITS

 receive

0.74

 Terms

0.73

 indemn

0.71

 abide

0.71

 Privacy

0.69

 Etsy

0.62

 cookies

0.62

 receiving

0.61

 Turtle

0.60

 confidentiality

0.60

Activations Density 0.024%

terms related to agreeing to privacy policies and receiving updates or offers

No Comments

No Known Activations