INDEX

Explanations

references to Western locations

oai_token-act-pair · gpt-3.5-turbo

references to "Western" entities or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

displayText

-0.86

staking

-0.80

ARP

-0.75

channelAvailability

-0.71

è¦ļéĨĴ

-0.70

UNCH

-0.69

INA

-0.68

LEASE

-0.68

inventoryQuantity

-0.68

gger

-0.67

POSITIVE LOGITS

 Hemisphere

0.96

 Union

0.88

 Sahara

0.85

 Bulldogs

0.81

 civilization

0.79

 Civilization

0.77

 Railway

0.77

ization

0.76

 democracies

0.75

 Brotherhood

0.75

Activations Density 0.012%

references to Western locations

references to "Western" entities or concepts

No Comments

No Known Activations