INDEX

Explanations

mentions of mental health-related terms or concepts

oai_token-act-pair · gpt-3.5-turbo

references to mental health topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

advertisement

-0.80

ICA

-0.78

IRD

-0.71

oulos

-0.71

ded

-0.69

aday

-0.69

gger

-0.68

ARDS

-0.68

rb

-0.67

ELS

-0.66

POSITIVE LOGITS

 faculties

0.98

 illness

0.93

 disorders

0.89

 defic

0.87

 disorder

0.84

itary

0.82

 wellbeing

0.81

izing

0.80

ising

0.80

 retard

0.79

Activations Density 0.012%

mentions of mental health-related terms or concepts

references to mental health topics

No Comments

No Known Activations