INDEX

Explanations

sequences and patterns involving variables, especially in mathematical contexts

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

 disambiguazione

-0.40

⚐

-0.40

 Ubicación

-0.38

HasAnnotation

-0.37

hoeddwyd

-0.36

 aplica

-0.35

 toepassing

-0.34

expandindo

-0.34

 encanto

-0.33

 aplicada

-0.33

POSITIVE LOGITS

 CreateTagHelper

0.53

isContained

0.49

UnusedPrivate

0.48

parsedMessage

0.48

apimachinery

0.47

IsContent

0.46

σιμοποι

0.46

UnitTesting

0.45

ztal

0.45

PyExc

0.44

Activations Density 0.053%

sequences and patterns involving variables, especially in mathematical contexts

No Comments

No Known Activations