INDEX

Explanations

phrases related to instructions or step-by-step guidance

oai_token-act-pair · gpt-3.5-turbo

various types of events and significant activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 contingency

-0.73

 recess

-0.73

 rede

-0.70

 presumptive

-0.69

 vacated

-0.68

 legislative

-0.67

 punitive

-0.67

 visitation

-0.66

 interim

-0.66

 UNCLASSIFIED

-0.65

POSITIVE LOGITS

 Featuring

1.07

 Includes

1.03

 curated

1.02

 Explore

0.99

 Learn

0.92

Includes

0.90

Features

0.86

 Including

0.86

 âĢº

0.86

 Contains

0.85

Activations Density 0.727%

phrases related to instructions or step-by-step guidance

various types of events and significant activities

No Comments

No Known Activations