INDEX

Explanations

phrases that express changes in belief or improved understanding over time

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 Brahmin

-0.64

 Administrativna

-0.64

 Parthen

-0.63

 gynhyrchwyd

-0.60

 intptr

-0.60

цездатний

-0.59

unzel

-0.58

 sacrament

-0.58

 plexus

-0.57

 Wikispecies

-0.56

POSITIVE LOGITS

 recently

0.61

LATEST

0.59

 lately

0.58

 Recently

0.58

 trend

0.57

hoeddwyd

0.55

 Recent

0.55

recently

0.55

 Lately

0.55

 recente

0.54

Activations Density 0.892%

phrases that express changes in belief or improved understanding over time

No Comments

No Known Activations