INDEX

Explanations

mentions or discussions related to sexual assault

oai_token-act-pair · gpt-3.5-turbo

references to the topic of rape

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Solitaire

-0.81

quickShipAvailable

-0.80

uchin

-0.79

ormons

-0.75

BLIC

-0.70

peror

-0.67

orsi

-0.67

ocular

-0.64

worldly

-0.63

rious

-0.63

POSITIVE LOGITS

 rape

1.13

 Rape

1.01

quez

0.90

 raped

0.87

 victim

0.87

rape

0.86

 rapes

0.86

 victims

0.85

 Victim

0.84

Sexual

0.82

Activations Density 0.038%

mentions or discussions related to sexual assault

references to the topic of rape

No Comments

No Known Activations