INDEX

Explanations

words related to discipline and control

oai_token-act-pair · gpt-3.5-turbo

mentions of discipline in various contexts

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

IFrame

Link

Negative Logits

istle

-0.71

swick

-0.69

DonaldTrump

-0.68

worth

-0.67

ophobic

-0.65

Phones

-0.65

stanbul

-0.63

izen

-0.63

 Citizens

-0.63

isp

-0.62

POSITIVE LOGITS

 discipline

1.02

cipline

0.87

ciplinary

0.86

 discipl

0.83

estinal

0.81

 disciplines

0.79

 disciplined

0.79

utical

0.78

cipl

0.78

 srfAttach

0.77

Activations Density 0.027%

words related to discipline and control

mentions of discipline in various contexts

No Known Activations