INDEX

Explanations

incidents involving injuries or harmful events

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

White House, Moore, Wright

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

featureID

-0.72

 كومونز

-0.68

 betweenstory

-0.62

Vidite

-0.59

afficheront

-0.58

⤹

-0.58

AccessorTable

-0.56

 surla

-0.56

parsedMessage

-0.56

 الحره

-0.53

POSITIVE LOGITS

 himself

0.41

 dirinya

0.41

 незавершена

0.40

himself

0.39

 herself

0.39

 Aseg

0.38

 rép

0.37

 terke

0.37

 palsu

0.36

SEGUIR

0.34

Activations Density 0.068%

incidents involving injuries or harmful events

White House, Moore, Wright

No Comments

No Known Activations