INDEX

Explanations

phrases related to privacy, private individuals or entities

oai_token-act-pair · gpt-3.5-turbo

references to private entities or activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

orthy

-0.84

wolves

-0.79

amaz

-0.75

tto

-0.72

————————

-0.69

andise

-0.66

older

-0.66

EFF

-0.66

addons

-0.65

azi

-0.64

POSITIVE LOGITS

 sector

1.19

 equity

0.99

sector

0.94

 jets

0.91

 investigator

0.91

 Sector

0.82

jet

0.81

 residences

0.80

 sphere

0.79

 detective

0.79

Activations Density 0.035%

phrases related to privacy, private individuals or entities

references to private entities or activities

No Comments

No Known Activations