INDEX

Explanations

specific named locations or businesses

oai_token-act-pair · gpt-3.5-turbo

possessive forms

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

Reviewer

-0.76

rait

-0.73

hari

-0.72

llor

-0.68

aneously

-0.67

arians

-0.67

ocument

-0.65

ENN

-0.64

ulkan

-0.64

mo

-0.63

POSITIVE LOGITS

 newest

0.96

 finest

0.93

own

0.90

 Revenge

0.84

 sake

0.83

 Syndrome

0.80

 signature

0.80

ullivan

0.79

 iconic

0.78

soDeliveryDate

0.77

Activations Density 0.163%

specific named locations or businesses

possessive forms

No Comments

No Known Activations