INDEX

Explanations

references to awards and recognitions within a specific context

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

awards and recognition

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

 autorytatywna

-0.70

 informée

-0.68

 ujednoznacz

-0.60

Ծանոթ

-0.57

évaluateur

-0.57

EndContext

-0.57

ArrowToggle

-0.53

 незавершена

-0.53

Vidite

-0.52

 JpaRepository

-0.52

POSITIVE LOGITS

 Awards

0.41

 Publishers

0.40

 awards

0.39

 Publisher

0.38

 rosette

0.36

 commend

0.36

omani

0.36

 magazine

0.36

 awarding

0.35

 manuf

0.35

Activations Density 0.017%

references to awards and recognitions within a specific context

awards and recognition

No Comments

No Known Activations