INDEX

Explanations

references to versioning and updates in software or data management contexts

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

version and revisions

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

ThroughAttribute

-0.90

 للمعارف

-0.76

riwal

-0.67

 InputDecoration

-0.65

httphttps

-0.63

 snippetHide

-0.62

ConstraintMaker

-0.61

Portail

-0.60

EndContext

-0.59

/\.

-0.59

POSITIVE LOGITS

runOnUiThread

0.37

 version

0.37

 update

0.36

 Version

0.33

 level

0.32

 mechanism

0.32

 updates

0.32

 participé

0.31

 Update

0.30

 twice

0.30

Activations Density 0.021%

references to versioning and updates in software or data management contexts

version and revisions

No Comments

No Known Activations