INDEX

Explanations

words related to violence and killing

oai_token-act-pair · gpt-3.5-turbo

instances of the word "kill" and its variations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ational

-0.79

BuyableInstoreAndOnline

-0.77

herty

-0.76

 Lich

-0.65

kj

-0.63

agall

-0.62

 Celest

-0.62

 Sparkle

-0.61

âĵĺ

-0.59

bles

-0.59

POSITIVE LOGITS

 spree

1.00

switch

0.98

mails

0.94

joy

0.90

houses

0.89

killer

0.89

blow

0.86

uminati

0.86

bird

0.84

 instinct

0.83

Activations Density 0.089%

words related to violence and killing

instances of the word "kill" and its variations

No Comments

No Known Activations