INDEX

Explanations

phrases indicating success or positive performance

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 plegable

-0.48

randomUUID

-0.42

annister

-0.41

 embraces

-0.40

 होना

-0.39

 أصبح

-0.39

 embrace

-0.39

TSCA

-0.39

-0.38

MediaStore

-0.38

POSITIVE LOGITS

CloseOperation

0.58

tagHelperRunner

0.57

 Done

0.55

 doGet

0.53

rungsseite

0.52

 doing

0.51

WaitGroup

0.51

 done

0.51

 doings

0.50

 DONE

0.49

Activations Density 0.013%

phrases indicating success or positive performance

No Comments

No Known Activations