INDEX

Explanations

phrases related to research methodologies and analyses

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

analysis and research evaluation

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

GEBURTSDATUM

-0.50

addCriterion

-0.50

Архівовано

-0.49

 Comprometido

-0.49

mybatisplus

-0.48

PostInfinity

-0.48

routeProvider

-0.47

 autorytatywna

-0.46

 الرياضيه

-0.46

 ThemeData

-0.43

POSITIVE LOGITS

wohnungen

0.36

ajur

0.35

Istorija

0.35

Redo

0.33

Kjelder

0.33

новременно

0.33

switchTo

0.33

LowerCase

0.33

Configs

0.33

ABLE

0.32

Activations Density 0.119%

phrases related to research methodologies and analyses

analysis and research evaluation

No Comments

No Known Activations