INDEX

Explanations

terms related to categorization and classification

oai_token-act-pair · gpt-3.5-turbo

terms related to categorization and classification processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

wives

-0.86

 Miracle

-0.75

 Reborn

-0.74

usa

-0.72

yre

-0.70

ixir

-0.69

sei

-0.69

 Roose

-0.68

EEK

-0.67

isk

-0.65

POSITIVE LOGITS

 categor

3.53

 classification

2.96

 classify

2.50

 Classification

2.47

 categorized

2.11

ategor

1.85

 classified

1.62

 categ

1.58

 categories

1.53

 category

1.51

Activations Density 0.022%

terms related to categorization and classification

terms related to categorization and classification processes

No Comments

No Known Activations