INDEX

Explanations

mentions of mental health related terms

oai_token-act-pair · gpt-3.5-turbo

references to mental health issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

advertisement

-0.82

gger

-0.75

ICA

-0.74

IRD

-0.73

ded

-0.70

aday

-0.70

eled

-0.68

ding

-0.66

oulos

-0.65

VEN

-0.64

POSITIVE LOGITS

 faculties

0.96

 disorders

0.91

 illness

0.86

 defic

0.83

 disorder

0.82

itary

0.80

 wellbeing

0.79

 retard

0.79

awar

0.79

 disabilities

0.79

Activations Density 0.011%

mentions of mental health related terms

references to mental health issues

No Comments

No Known Activations