INDEX

Explanations

references to having systems, policies, or guidelines in place

oai_token-act-pair · gpt-3.5-turbo

references to established policies or systems

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

initely

-0.83

jin

-0.76

yip

-0.74

cest

-0.73

incinn

-0.73

uddy

-0.70

zzy

-0.69

yssey

-0.69

ingly

-0.67

anche

-0.67

POSITIVE LOGITS

bos

0.90

 antioxid

0.79

holders

0.76

ascript

0.71

Ĥİ

0.71

 defences

0.69

holder

0.68

lie

0.66

å§«

0.64

,,,,,,,,

0.63

Activations Density 0.018%

references to having systems, policies, or guidelines in place

references to established policies or systems

No Comments

No Known Activations