INDEX

Explanations

instances where someone is determined to take action or do something

oai_token-act-pair · gpt-3.5-turbo

references to taking action or doing something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

aten

-0.70

uart

-0.66

eri

-0.64

bda

-0.64

opes

-0.63

mallow

-0.63

ockets

-0.62

ulia

-0.62

ebin

-0.61

 sshd

-0.61

POSITIVE LOGITS

 wrong

1.10

 differently

1.06

 stupid

1.05

wrong

1.04

 drastic

1.03

 proactive

0.99

 crazy

0.98

 rash

0.97

 risky

0.96

 unusual

0.94

Activations Density 0.059%

instances where someone is determined to take action or do something

references to taking action or doing something

No Comments

No Known Activations