INDEX

Explanations

terms related to capacity or ability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 structured

-0.55

 style

-0.51

 styled

-0.49

Ped

-0.49

 vision

-0.48

Nar

-0.48

Mig

-0.47

hept

-0.47

 design

-0.47

 Providing

-0.47

POSITIVE LOGITS

OutputStream

0.78

InputStream

0.76

genodigd

0.66

 mijne

0.64

 Ahnung

0.63

 hjär

0.62

 zijne

0.60

 unterkunft

0.57

 vägen

0.57

 חיצוניים

0.56

Activations Density 0.186%

terms related to capacity or ability

No Comments

No Known Activations