INDEX

Explanations

instances of the term "Fig" followed by a number

oai_token-act-pair · gpt-3.5-turbo

references to figures and illustrations

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 administ

-0.74

 conditioning

-0.70

 convict

-0.67

×Ĳ

-0.62

 Belarus

-0.61

etheless

-0.60

×ĳ

-0.60

 hearing

-0.59

ngth

-0.59

 precious

-0.58

POSITIVE LOGITS

ures

1.35

uration

1.31

ured

1.26

uring

1.24

urations

1.19

aro

1.18

uer

1.12

URE

1.08

uers

1.05

ues

1.03

Activations Density 0.026%

instances of the term "Fig" followed by a number

references to figures and illustrations

No Comments

No Known Activations