INDEX

Explanations

punctuation marks and apostrophes

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

 laikā

-0.47

 χρόν

-0.46

 Könige

-0.44

esistenza

-0.42

 utiles

-0.42

 tarko

-0.41

 Unterricht

-0.41

 útiles

-0.41

 aikana

-0.41

 spreken

-0.41

POSITIVE LOGITS

 ***!

0.61

 Paglinawan

0.60

RegistryLite

0.57

 PyLong

0.57

 gynhyrchwyd

0.57

 TMPro

0.57

StructEnd

0.57

setVerticalGroup

0.56

 дописавши

0.54

=$?

0.53

Activations Density 0.090%

punctuation marks and apostrophes

No Comments

No Known Activations