INDEX

Explanations

legal and technical terms related to documentation and data processing

oai_token-act-pair · gpt-3.5-turbo

references to technical terms, processes, and structures in documentation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Yep

-0.73

Had

-0.73

Damn

-0.68

etheless

-0.67

Angelo

-0.65

Fuck

-0.63

ZI

-0.61

 Damn

-0.60

Reviewed

-0.60

Got

-0.59

POSITIVE LOGITS

 consists

1.64

 ensures

1.45

 lasts

1.41

 varies

1.36

 differs

1.33

 allows

1.33

 involves

1.33

 determines

1.32

 includes

1.32

 comprises

1.29

Activations Density 0.422%

legal and technical terms related to documentation and data processing

references to technical terms, processes, and structures in documentation

No Comments

No Known Activations