INDEX

Explanations

terms related to automation or automated processes

oai_token-act-pair · gpt-3.5-turbo

references to automated systems and processes

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

dn

-0.87

hips

-0.82

nee

-0.81

hov

-0.78

ings

-0.74

adr

-0.73

lust

-0.73

thur

-0.72

Magikarp

-0.72

uin

-0.72

POSITIVE LOGITS

 automated

1.19

 automation

1.12

 automate

1.01

 autom

0.93

 captcha

0.90

 systems

0.86

 guiActiveUn

0.82

 bots

0.81

 humming

0.78

 cheating

0.77

Activations Density 0.016%

terms related to automation or automated processes

references to automated systems and processes

No Known Activations