INDEX

Explanations

elements and symbols related to mathematical structures and expressions

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

parameters and code

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

 snippetHide

-0.58

 يتيمه

-0.52

 препратки

-0.51

odetic

-0.50

medriver

-0.49

culable

-0.47

 Paglinawan

-0.47

cipated

-0.46

♮

-0.46

POSITIVE LOGITS

ParallelGroup

0.42

DrawerToggle

0.37

 Refer

0.36

 ब्रेकडाउन

0.34

 autorytatywna

0.33

sendRedirect

0.33

Cringe

0.32

 PyLong

0.32

 burnout

0.32

endphp

0.32

Activations Density 0.002%

elements and symbols related to mathematical structures and expressions

parameters and code

No Comments

No Known Activations