INDEX

Explanations

common words that indicate relationships or comparisons, particularly involving "to" and "as."

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

as far as, in terms of, when it comes to

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

 soudain

-0.43

 bouteille

-0.38

hideLoading

-0.36

 boneca

-0.35

 goutte

-0.34

 préc

-0.32

 déposer

-0.31

 sœur

-0.31

 garrafa

-0.30

 Heiligen

-0.30

POSITIVE LOGITS

 بشأن

0.63

RTEX

0.62

 assessing

0.62

xtext

0.61

__*/

0.61

}{*}{

0.60

 мәкал

0.59

 щодо

0.59

XMLSchema

0.59

 kaarangay

0.59

Activations Density 0.017%

common words that indicate relationships or comparisons, particularly involving "to" and "as."

as far as, in terms of, when it comes to

No Comments

No Known Activations