INDEX

Explanations

phrases related to figures and diagrams

oai_token-act-pair · gpt-3.5-turbo

references to figures or visual aids within the text

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

TPPStreamerBot

-0.71

 Sussex

-0.71

 Kimmel

-0.68

ngth

-0.66

 Barclays

-0.63

×©

-0.63

 Passenger

-0.62

UGH

-0.62

ITY

-0.62

×ª

-0.61

POSITIVE LOGITS

uring

1.20

ured

1.19

uration

1.18

ures

1.15

aro

1.06

urations

1.05

uer

0.96

uers

0.92

ue

0.91

ging

0.91

Activations Density 0.015%

phrases related to figures and diagrams

references to figures or visual aids within the text

No Comments

No Known Activations