INDEX

Explanations

terms related to data handling and file processing

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

technical variable assignments

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

feier

-0.37

プローチ

-0.34

 PyTuple

-0.34

 Acuerdo

-0.32

 değil

-0.32

 Vertrags

-0.31

sau

-0.31

})-

-0.30

to

-0.30

 bütün

-0.29

POSITIVE LOGITS

endphp

0.78

IANGLES

0.74

expandindo

0.73

 GenerationType

0.71

 ProtoMessage

0.69

parsedMessage

0.69

 EconPapers

0.65

styleable

0.64

bootstrapcdn

0.63

 Audiodateien

0.63

Activations Density 0.031%

terms related to data handling and file processing

technical variable assignments

No Comments

No Known Activations