INDEX

Explanations

numeric values and codes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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



-0.69

 Anfitrión

-0.52

 autorytatywna

-0.51

izoen

-0.47

 литров

-0.46

 beams

-0.46

 terms

-0.46

Datuak

-0.45

########.

-0.45

 рік

-0.45

POSITIVE LOGITS

⅘

0.43

Côte

0.41

modelBuilder

0.41

modium

0.40

 craindre

0.40

Distinguished

0.39

webElementGuid

0.39

 XNUMX

0.39

Superior

0.38

 prochaines

0.38

Activations Density 0.157%

numeric values and codes

No Comments

No Known Activations