INDEX

Explanations

patterns of whitespace or formatting in the text

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

technical data or legal citations

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

 EconPapers

-0.68

 Roskov

-0.53

 transfieras

-0.51

prüche

-0.49

 nahilalakip

-0.49

 nakalista

-0.47

cyd

-0.46

atimes

-0.43

endregion

-0.43

mybatisplus

-0.43

POSITIVE LOGITS

 Anſ

0.43

ſelf

0.41

 anſ

0.41

 ſta

0.41

 houſe

0.40

 preſ

0.39

Preparazione

0.39

 nonUne

0.38

AnimationsModule

0.38

 Photocase

0.37

Activations Density 0.000%

patterns of whitespace or formatting in the text

technical data or legal citations

No Comments

No Known Activations