INDEX

Explanations

phrases related to rules, regulations, and instructions

oai_token-act-pair · gpt-3.5-turbo

instances of slashes or dividers in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

inois

-0.69

 stalking

-0.68

 creep

-0.68

ļéĨĴ

-0.67

atche

-0.67

 sooner

-0.66

 imperson

-0.66

ĸļ

-0.66

Obi

-0.65

DeV

-0.65

POSITIVE LOGITS

etc

1.25

pmwiki

1.05

whatever

1.03

AFP

1.02

Getty

1.02

CN

0.99

sites

0.95

Flickr

0.94

Activations Density 0.103%

phrases related to rules, regulations, and instructions

instances of slashes or dividers in the text

No Comments

No Known Activations