INDEX

Explanations

specific time references and intervals

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

̈́

-0.48

Pyx

-0.46

StoreMessageInfo

-0.42

modelBuilder

-0.41

 Morning

-0.40

 oprot

-0.40

Morning

-0.40

 Yearbook

-0.38

twimg

-0.38

viaf

-0.38

POSITIVE LOGITS

 weeks

0.84

weeks

0.72

Weeks

0.68

 settimane

0.67

 Weeks

0.66

 semaines

0.64

 weken

0.64

 week

0.64

 month

0.64

 WEEKS

0.63

Activations Density 0.198%

specific time references and intervals

No Comments

No Known Activations