INDEX

Explanations

references to official approvals and regulatory actions

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

permission granted

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

IfExists

-0.31

 Kupfer

-0.29

upsi

-0.29

 vergüenza

-0.29

CrossRef

-0.28

 peças

-0.28

 Norbert

-0.27

Jo

-0.26

jo

-0.26

 presenting

-0.26

POSITIVE LOGITS

 approved

0.79

 permitted

0.78

 allowed

0.78

 approves

0.77

 approve

0.76

 approval

0.75

approved

0.73

 Allowed

0.73

 permits

0.73

 Autorisations

0.72

Activations Density 0.051%

references to official approvals and regulatory actions

permission granted

No Comments

No Known Activations