INDEX

Explanations

mentions of new developments or changes in projects or products

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

coming this fall

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

nė

-0.44

 eventual

-0.42

seguida

-0.40

 eventualmente

-0.40

ulman

-0.39

 πάν

-0.38

LIE

-0.38

 instant

-0.38

 rodillas

-0.37

 eventually

-0.37

POSITIVE LOGITS

ngdoc

0.57

 AssemblyTitle

0.54

 الرياضيه

0.53

StoryboardSegue

0.51

 ModelExpression

0.51

VersionUID

0.50



0.50

setVerticalGroup

0.49

出版年

0.49

 NSCoder

0.49

Activations Density 0.037%

mentions of new developments or changes in projects or products

coming this fall

No Comments

No Known Activations