INDEX

Explanations

references to prequels or sequels in narratives

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

achella

-0.54

 onPostExecute

-0.46

ſem

-0.44

intellij

-0.43

ſo

-0.43

ſtra

-0.42

 Alegre

-0.41

aulted

-0.41

llery

-0.41

ambién

-0.40

POSITIVE LOGITS

tonode

0.61

EndGlobalSection

0.49

 prequel

0.49

DockStyle

0.48

Period

0.42

 пред

0.41

Diwedd

0.39

 nonUne

0.37

keyColumn

0.37

WithIOException

0.36

Activations Density 0.012%

references to prequels or sequels in narratives

No Comments

No Known Activations