INDEX

Explanations

usernames or aliases

oai_token-act-pair · gpt-3.5-turbo

terms related to usernames and aliases

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

Reviewed

-0.71

aud

-0.69

urb

-0.68

oons

-0.67

icult

-0.67

itsch

-0.67

OTT

-0.67

aith

-0.66

akh

-0.64

yrinth

-0.64

POSITIVE LOGITS

 username

0.94

 Username

0.92

 aliases

0.88

username

0.87

 alias

0.85

 belonging

0.81

 persona

0.79

 Generator

0.78

ccess

0.73

 surname

0.73

Activations Density 0.044%

usernames or aliases

terms related to usernames and aliases

No Comments

No Known Activations