INDEX

Explanations

terms related to approval and influence, particularly in contexts involving ratings or rankings

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

based on technical terms or non-English words

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

 diken

-0.29

 policiales

-0.29

dé

-0.29

 Pugh

-0.29

ayr

-0.28

 genieten

-0.28

NSYLVANIA

-0.28

 desses

-0.27

 abbraccio

-0.27

 reposer

-0.26

POSITIVE LOGITS

XmlAccessType

0.64

 Мексичка

0.63

 defaultstate

0.61

 autorytatywna

0.61

rungsseite

0.60

 MainAxisSize

0.60

afficheront

0.60

Autoritní

0.59

 wireType

0.56

 للاسماء

0.55

Activations Density 0.280%

terms related to approval and influence, particularly in contexts involving ratings or rankings

based on technical terms or non-English words

No Comments

No Known Activations