INDEX

Explanations

references to structural elements and attributes in code or data formats

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits



-0.47

Jeografia

-0.45

zzleHttp

-0.45

Yak

-0.44

prefixer

-0.43

󠁮

-0.42

 EconPapers

-0.42

lap

-0.42

SneakyThrows

-0.42

fitriones

-0.42

POSITIVE LOGITS

новништво

0.51

deutig

0.45

 administratif

0.40

AZINE

0.38

einander

0.37

 polaire

0.37

 localObject

0.37

 dziecko

0.37

 meille

0.36

舺

0.36

Activations Density 1.146%

references to structural elements and attributes in code or data formats

No Comments

No Known Activations