INDEX

Explanations

metrics related to diagnostic accuracy and performance evaluation

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

Metrics of model performance

np_acts-logits-general · gemini-2.0-flash

evaluation metrics

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 alber

-0.33

Lit

-0.30

Zum

-0.29

Gru

-0.28

 club

-0.28

Tip

-0.28

Bronnen

-0.28

en

-0.28

 tils

-0.27

:+:

-0.27

POSITIVE LOGITS

 accuracy

0.87

 Accuracy

0.73

accuracy

0.69

 TestBed

0.65

elemField

0.63

 ainfi

0.62

Accuracy

0.60

 desempenho

0.59

featureID

0.58

 performance

0.57

Activations Density 0.995%

metrics related to diagnostic accuracy and performance evaluation

Metrics of model performance

evaluation metrics

No Comments

No Known Activations

metrics related to diagnostic accuracy and performance evaluation

Metrics of model performance

evaluation metrics

No Comments

No Known Activations