INDEX

Explanations

mathematical expressions and relationships involving variables and parameters

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

original and enhanced

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

Спасылкі

-0.50

 épaules

-0.46

AndEndTag

-0.44

 Usaha

-0.44

 travaillons

-0.43

ziasztok

-0.42

úsica

-0.42

 velkommen

-0.41

ſelf

-0.41

 määrä

-0.41

POSITIVE LOGITS

(!__

0.50

 AssemblyCulture

0.47

 Италијани

0.42

 Taktlose

0.40

<!--[

0.39

 <=",

0.39

matchCondition

0.39

baomidou

0.38

 Lond

0.37

Архівовано

0.35

Activations Density 0.125%

mathematical expressions and relationships involving variables and parameters

original and enhanced

No Comments

No Known Activations