INDEX

Explanations

positive qualities or high levels of significance in text

oai_token-act-pair · gpt-3.5-turbo

descriptors of quality and speed

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Maze

-0.69

odder

-0.59

adelphia

-0.58

 Nile

-0.58

 Ariel

-0.58

Ki

-0.58

 Insight

-0.57

 Nath

-0.57

 Jude

-0.57

EE

-0.56

POSITIVE LOGITS

ractive

0.79

(>

0.76

tarian

0.69

nered

0.66

xual

0.63

 sexism

0.62

auld

0.62

uilt

0.62

 compliments

0.60

ifiable

0.60

Activations Density 0.321%

positive qualities or high levels of significance in text

descriptors of quality and speed

No Comments

No Known Activations