INDEX

Explanations

terms related to cults

oai_token-act-pair · gpt-3.5-turbo

variations of the word "ult."

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

 Blaz

-0.71

BY

-0.70

 scratch

-0.67

CHR

-0.66

 excuse

-0.65

strings

-0.65

hold

-0.61

 rough

-0.59

forth

-0.58

 spare

-0.58

POSITIVE LOGITS

ult

1.23

imately

1.13

iple

1.12

urally

1.07

iful

1.04

icultural

1.04

raviolet

1.04

uration

1.03

icult

1.02

anamo

1.01

Activations Density 0.005%

terms related to cults

variations of the word "ult."

No Comments

No Known Activations