INDEX

Explanations

phrases related to victims of various forms of harm or injustice

oai_token-act-pair · gpt-3.5-turbo

mentions of victims in various contexts

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

leaf

-0.71

ATIONAL

-0.69

66666666

-0.66

GGGG

-0.66

 monarchy

-0.65

 Plat

-0.65

CLASSIFIED

-0.64

obar

-0.63

Sau

-0.62

arp

-0.61

POSITIVE LOGITS

 Victims

1.17

 victims

1.13

 Survivors

0.94

 Victim

0.94

 victim

0.85

 survivors

0.84

 inflicted

0.83

 victimized

0.83

vict

0.79

 Survivor

0.76

Activations Density 0.013%

phrases related to victims of various forms of harm or injustice

mentions of victims in various contexts

No Comments

No Known Activations