INDEX

Explanations

terms related to violent acts or crimes

oai_token-act-pair · gpt-3.5-turbo

references to violent acts, specifically killings and related terminology

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Cola

-0.77

BN

-0.70

Stud

-0.69

 Plex

-0.68

aque

-0.68

amina

-0.67

BuyableInstoreAndOnline

-0.64

inion

-0.64

air

-0.63

Dub

-0.62

POSITIVE LOGITS

 shootings

1.07

 spree

1.04

 killings

1.02

hooting

0.93

 murders

0.91

mith

0.88

 homicides

0.87

poons

0.87

 slaying

0.87

 shooting

0.86

Activations Density 0.020%

terms related to violent acts or crimes

references to violent acts, specifically killings and related terminology

No Comments

No Known Activations