INDEX

Explanations

words related to criminal activities or legal terms

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "crime."

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

zl

-0.72

uran

-0.72

estine

-0.70

ersed

-0.69

artney

-0.69

TPPStreamerBot

-0.65

pass

-0.64

}}}

-0.64

achev

-0.63

 Pixie

-0.62

POSITIVE LOGITS

 spree

1.02

 crime

1.00

 Crime

0.91

 crimes

0.86

 homicide

0.85

 synd

0.84

fighting

0.84

 homicides

0.81

 Victim

0.81

 unfocusedRange

0.80

Activations Density 0.017%

words related to criminal activities or legal terms

occurrences of the word "crime."

No Comments

No Known Activations