INDEX

Explanations

phrases referring to specific things or concepts mentioned in the document

oai_token-act-pair · gpt-3.5-turbo

phrases that highlight key elements or components in a list or description

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

conn

-0.66

 alloc

-0.65

tern

-0.60

 develops

-0.57

lab

-0.57

ls

-0.56

bil

-0.56

 unsuccessfully

-0.56

gain

-0.56

lete

-0.55

POSITIVE LOGITS

hett

0.65

Ķ

0.63

bably

0.63

ÑĮ

0.62

yout

0.60

enta

0.60

cient

0.59

CRIPTION

0.59

omething

0.59

ultimate

0.58

Activations Density 0.202%

phrases referring to specific things or concepts mentioned in the document

phrases that highlight key elements or components in a list or description

No Comments

No Known Activations