INDEX

Explanations

references to individuals and their relationships to distinct entities or concepts, particularly in a historical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 overarching

-0.60

 moniker

-0.60

 timeline

-0.58

 sparkly

-0.57

 hallways

-0.57

 vibrant

-0.56

 heist

-0.54

 standalone

-0.54

 nuanced

-0.54

 trif

-0.54

POSITIVE LOGITS

 muß

0.57

 läßt

0.56

 groote

0.48

 foregoing

0.46

 biß

0.45

 andern

0.45

 daß

0.45

 varias

0.44

 übri

0.44

 Schluß

0.44

Activations Density 1.013%

references to individuals and their relationships to distinct entities or concepts, particularly in a historical context

No Comments

No Known Activations