INDEX

Explanations

references to legal terms and concepts, particularly related to prejudice and alternative arguments

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

prejudice, manifestation, walking, natural

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

Superficie

-0.38

บรร

-0.37

On

-0.35

 total

-0.35

In

-0.34

The

-0.33

epam

-0.33

 aikaa

-0.32

Call

-0.32

itsi

-0.31

POSITIVE LOGITS

 للاسماء

0.83

 незавершена

0.78

StoryboardSegue

0.72

tangentMode

0.68

 informée

0.66

GOTREF

0.66

OGND

0.65

EndProject

0.64

httphttps

0.63

Hentet

0.63

Activations Density 0.074%

references to legal terms and concepts, particularly related to prejudice and alternative arguments

prejudice, manifestation, walking, natural

No Comments

No Known Activations