INDEX

Explanations

direct speech or quotations

oai_token-act-pair · gpt-3.5-turbo

statements related to governmental reports or announcements

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

 taxp

-0.65

tumblr

-0.65

soDeliveryDate

-0.64

 stub

-0.62

nig

-0.59

 pound

-0.57

 perfected

-0.54

 Rolls

-0.54

 semblance

-0.54

 lasts

-0.54

POSITIVE LOGITS

agascar

0.66

nces

0.65

enegger

0.64

adena

0.62

yrinth

0.61

inguished

0.61

today

0.61

etheless

0.61

zai

0.61

uckland

0.61

Activations Density 0.646%

direct speech or quotations

statements related to governmental reports or announcements

No Comments

No Known Activations