INDEX

Explanations

phrases related to negative actions and behaviors towards others

oai_token-act-pair · gpt-3.5-turbo

conjunctions and phrases indicating an ongoing relationship or connection between ideas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

Yao

-0.71

 Shrine

-0.64

 Animation

-0.63

 Romans

-0.62

Goo

-0.62

 Colts

-0.62

NRL

-0.61

Sox

-0.60

 Dungeons

-0.59

 Shots

-0.59

POSITIVE LOGITS

rogen

1.06

rogens

1.04

 distribute

0.85

 punish

0.83

 rehabilit

0.83

 humili

0.80

 analyse

0.79

rew

0.79

 manipulate

0.78

 dispose

0.77

Activations Density 0.131%

phrases related to negative actions and behaviors towards others

conjunctions and phrases indicating an ongoing relationship or connection between ideas

No Comments

No Known Activations