INDEX

Explanations

mention of the word "torture"

oai_token-act-pair · gpt-3.5-turbo

terms related to torture

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

 Darkness

-0.78

 magnification

-0.75

 Prospect

-0.69

ACP

-0.67

 Silk

-0.66

ECB

-0.65

 donor

-0.64

 brightest

-0.64

 Flavoring

-0.63

lihood

-0.63

POSITIVE LOGITS

urous

1.45

oise

1.38

uous

1.06

uring

1.04

illas

1.01

ured

0.98

eur

0.98

oled

0.96

imer

0.94

ificate

0.93

Activations Density 0.012%

mention of the word "torture"

terms related to torture

No Comments

No Known Activations