INDEX

Explanations

terms related to improvement and optimization in performance metrics

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

coaches and skills

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 Administrative

-0.33

Administrative

-0.31

)];

-0.31

 confe

-0.31

 administrative

-0.31

onames

-0.30

administrative

-0.29

 cryptographic

-0.28

 Comun

-0.28

sender

-0.28

POSITIVE LOGITS

tagext

0.70

featureID

0.70

 resourceCulture

0.68

<?

0.67

IntoConstraints

0.66

StructEnd

0.65

PropertyChanging

0.65

 coaches

0.65

getColumnIndex

0.62

 Coaches

0.62

Activations Density 0.018%

terms related to improvement and optimization in performance metrics

coaches and skills

No Comments

No Known Activations