INDEX

Explanations

programming-related boolean values and keywords

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

ніципалі

-0.67

 utafitiHapana

-0.59

findpost

-0.56

GEBURTSDATUM

-0.46

expandindo

-0.44

">—

-0.43

')):

-0.43

fromnode

-0.41

adaptiveStyles

-0.41

 PeEnEo

-0.40

POSITIVE LOGITS

leneck

0.56

 supérieurs

0.52

 Extended

0.48

 purpoſe

0.48

BagConstraints

0.47

 alimentaires

0.47

 eşofman

0.47

 sanitarias

0.46

 económicas

0.46

 complejidad

0.45

Activations Density 0.000%

programming-related boolean values and keywords

No Comments

No Known Activations