INDEX

Explanations

mentions of the country India

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Ind" or terms associated with it in various contexts

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

sonian

-0.89

 compr

-0.87

EStream

-0.86

avorite

-0.76

 litter

-0.74

TION

-0.70

Ĥİ

-0.68

RTX

-0.65

collection

-0.65

xit

-0.64

POSITIVE LOGITS

irect

1.31

ividually

1.29

icative

1.23

ebted

1.21

icator

1.19

icates

1.12

ications

1.11

icators

1.09

ifferent

1.09

enture

1.07

Activations Density 0.015%

mentions of the country India

mentions of the word "Ind" or terms associated with it in various contexts

No Comments

No Known Activations