INDEX

Explanations

underscore-prefixed terms, typically representing private or protected variables or methods in programming code

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

Declaration of Helsinki

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

 gwiazd

-0.45

 verbonden

-0.44

 extranjero

-0.44

 Belast

-0.44

 bebas

-0.44

ωση

-0.44

 either

-0.43

ActionCreators

-0.43

on

-0.43

 Jerusalén

-0.43

POSITIVE LOGITS

Autoritní

0.68

PreferredItem

0.64

 FontWeight

0.61

 autorytatywna

0.56

 ویکی‌پدی

0.54

 Paglinawan

0.51

AutoScaleMode

0.47

Bibliograf

0.46

_{-\

0.46

hematical

0.45

Activations Density 0.017%

underscore-prefixed terms, typically representing private or protected variables or methods in programming code

Declaration of Helsinki

No Comments

No Known Activations