INDEX

Explanations

words related to moral excellence or principles

oai_token-act-pair · gpt-3.5-turbo

references to the concept of virtue

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ockets

-0.76

WER

-0.71

 Niet

-0.67

mable

-0.66

ears

-0.65

è¦ļéĨĴ

-0.64

 Crosby

-0.63

STER

-0.63

oval

-0.62

grown

-0.62

POSITIVE LOGITS

 virtue

0.96

 signalling

0.91

 Virtue

0.86

ocracy

0.85

 signaling

0.82

uous

0.78

iencies

0.76

ocratic

0.74

ienne

0.71

cean

0.71

Activations Density 0.015%

words related to moral excellence or principles

references to the concept of virtue

No Comments

No Known Activations