INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

mentions of specific entities or abbreviations, particularly related to institutions or organizations

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

IZE

-0.70

perate

-0.67

igans

-0.66

Construct

-0.63

IAN

-0.62

ãĥī

-0.61

 Monaco

-0.61

 Nadu

-0.59

utenant

-0.58

 ORIG

-0.58

POSITIVE LOGITS

ruary

1.09

acteria

1.05

ilib

1.05

razil

1.05

iotics

1.01

axter

1.00

odies

0.97

yssey

0.96

ishop

0.96

amboo

0.95

Activations Density 0.034%

proper nouns

mentions of specific entities or abbreviations, particularly related to institutions or organizations

No Comments

No Known Activations