INDEX

Explanations

terms related to well-being and happiness

oai_token-act-pair · gpt-3.5-turbo

references to well-being and its associated practices

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

•layer 3 - joy

No Comments

Negative Logits

 tiss

-0.65

ulla

-0.64

oslov

-0.63

 racks

-0.63

 Lilly

-0.62

NetMessage

-0.61

 rack

-0.61

ERY

-0.60

Ey

-0.58

PLA

-0.58

POSITIVE LOGITS

 able

0.76

Angelo

0.74

packs

0.71

 mindful

0.69

 bearer

0.69

ipolar

0.68

held

0.67

Unit

0.65

iverse

0.65

 endowed

0.64

Activations Density 0.038%

terms related to well-being and happiness

references to well-being and its associated practices

No Comments

No Known Activations