INDEX

Explanations

references to 'TS'

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "TS" along with related numerical or categorization terms

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

liest

-0.81

ston

-0.70

 joke

-0.69

ously

-0.68

 laundry

-0.67

stein

-0.64

ances

-0.63

hold

-0.63

 Cosby

-0.63

esville

-0.62

POSITIVE LOGITS

weet

1.11

ullivan

1.00

BUR

0.99

ypes

0.91

omething

0.91

ocial

0.91

ierra

0.89

ICLE

0.87

WER

0.86

ystem

0.85

Activations Density 0.015%

references to 'TS'

mentions of the term "TS" along with related numerical or categorization terms

No Comments

No Known Activations