INDEX

Explanations

terms related to data analysis and monitoring processes

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

action or status update

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

Personensuche

-0.60

ckså

-0.53

UrlResolution

-0.47

SOUNDBITE

-0.47

 فريبيس

-0.47

 LoginComponent

-0.46

IBarButtonItem

-0.45

triple

-0.44

ýt

-0.43

Organisateur

-0.43

POSITIVE LOGITS

 rebajas

0.35

ⓧ

0.35

 köt

0.35

قایناق‌لار

0.35

 retr

0.34

setcounter

0.33

facade

0.33

LOYEES

0.32

➯

0.32

charest

0.31

Activations Density 0.276%

terms related to data analysis and monitoring processes

action or status update

No Comments

No Known Activations