INDEX

Explanations

references to positions of authority, office, and responsibilities

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

holding a position or office

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

 MainAxisSize

-0.63

 jsPsych

-0.60

EndGlobalSection

-0.57

دانشنامهٔ

-0.57

 CreateTagHelper

-0.51

adaptiveStyles

-0.50

EndProject

-0.49

IntoConstraints

-0.48

TagMode

-0.48

 おっぱい

-0.47

POSITIVE LOGITS

 position

0.52

 leadership

0.41

Position

0.39

 appointment

0.38

apimachinery

0.36

position

0.35

 houſe

0.35

 Position

0.34

 office

0.34

 positions

0.34

Activations Density 0.035%

references to positions of authority, office, and responsibilities

holding a position or office

No Comments

No Known Activations