INDEX

Explanations

phrases related to potential harm or danger

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the concept of risk and danger to lives or communities

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Britt

-0.69

tein

-0.68

lins

-0.68

anche

-0.66

ION

-0.66

Nap

-0.65

encer

-0.64

frey

-0.63

NC

-0.62

 Kinnikuman

-0.62

POSITIVE LOGITS

taker

0.77

taking

0.76

»Ĵ

0.74

groups

0.70

 situations

0.69

ħĭ

0.68

xual

0.67

ailability

0.66

 stewards

0.65

orses

0.65

Activations Density 0.012%

phrases related to potential harm or danger

phrases indicating the concept of risk and danger to lives or communities

No Comments

No Known Activations