INDEX

Explanations

references to historical figures and events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Vue

-0.55

 Aene

-0.55

vue

-0.55

 Etrus

-0.53

Vue

-0.52

 Bajaj

-0.51

UVWXYZ

-0.51

 Prussia

-0.49

 ſte

-0.49

 Canberra

-0.49

POSITIVE LOGITS

 escolas

0.44

 escolares

0.43

 camada

0.42

 combatir

0.40

 JpaRepository

0.40

 fontes

0.39

 scolaires

0.39

 versões

0.38

 cráneo

0.38

 bancada

0.37

Activations Density 0.123%

references to historical figures and events

No Comments

No Known Activations