INDEX

Explanations

highlights related to specific nationalities

oai_token-act-pair · gpt-3.5-turbo

references to nationalities, particularly French, Chinese, and Japanese

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ctic

-0.87

odder

-0.84

ueller

-0.83

lier

-0.78

affles

-0.78

vable

-0.78

inery

-0.78

utherford

-0.77

izons

-0.77

idem

-0.77

POSITIVE LOGITS

 Nadu

0.98

 proverb

0.94

oslov

0.94

 nationals

0.84

 cuisine

0.84

 pronunciation

0.82

 immigrants

0.81

apolis

0.80

Yen

0.80

 immigrant

0.80

Activations Density 0.130%

highlights related to specific nationalities

references to nationalities, particularly French, Chinese, and Japanese

No Comments

No Known Activations