INDEX

Explanations

references to directories and folders in a technical context

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

"folder" or "chapter"

np_acts-logits-general · gemini-2.0-flash

folder or chapter

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Prest

-0.54

})`

-0.49

'],'

-0.47

forced

-0.47

 Barb

-0.45

 Forced

-0.45

<bos>

-0.44

-0.42

Forced

-0.42

compute

-0.41

POSITIVE LOGITS

Chapter

0.92

 Chapter

0.91

 chapter

0.91

 chapters

0.86

 Kapitel

0.86

 Chapters

0.84

Kapitel

0.83

Chapitre

0.79

Chapters

0.79

chapter

0.79

Activations Density 0.161%

references to directories and folders in a technical context

"folder" or "chapter"

folder or chapter

No Comments

No Known Activations