INDEX

Explanations

numerical values and their representations in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

AccessorTable

-0.47

 pinulongan

-0.42

 terciopelo

-0.41

PropertyChanging

-0.40

 referrerpolicy

-0.37

 desafíos

-0.36

GrantedAuthority

-0.36

+#+#

-0.35

 Feuchtigkeit

-0.34

 ensayos

-0.34

POSITIVE LOGITS

IsMutable

0.45

 contextLoads

0.44

Octa

0.44

Fol

0.44

 Chic

0.43

Fol

0.43

 Octave

0.42

 radial

0.41

 Doug

0.41

Chic

0.40

Activations Density 0.746%

numerical values and their representations in various contexts

No Comments

No Known Activations