INDEX

Explanations

instances of discourse markers or transitional phrases in text

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

 kasarigan

-0.62

Rptr

-0.58

CppMethod

-0.57

GEBURTSDATUM

-0.56

 pinulongan

-0.55

ioutil

-0.54

Бахар

-0.52

 autorytatywna

-0.52

WaitGroup

-0.52

__":

-0.51

POSITIVE LOGITS

 HFILL

0.35

 utafitiHapana

0.34

 boden

0.33

 esternos

0.32

andten

0.30

pexpr

0.30

 accesso

0.30

 postseason

0.29

gehend

0.29

 forfait

0.28

Activations Density 0.023%

instances of discourse markers or transitional phrases in text

No Comments

No Known Activations