INDEX

Explanations

terms related to various scientific and technical measurements or assessments

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

measured reported understanding

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

WriteTagHelper

-0.65

ByVersion

-0.63

transQ

-0.61

Tembelea

-0.59

Obrázky

-0.59

setDo

-0.58

 виправивши

-0.57

 فريبيس

-0.57

 protoimpl

-0.55

AddTagHelper

-0.54

POSITIVE LOGITS

 ambientales

0.34

一樣

0.33

 similarly

0.33

 ebenfalls

0.31

 likewise

0.31

 direta

0.31

 szint

0.31

 esportivo

0.31

编程

0.31

 ácido

0.31

Activations Density 0.299%

terms related to various scientific and technical measurements or assessments

measured reported understanding

No Comments

No Known Activations