INDEX

Explanations

references to specific TV shows and their characters

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

TV show characters and titles

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

 municipalité

-0.40

 Riesen

-0.34

UNICIP

-0.34

 Cowan

-0.34

 popolo

-0.34

 municipality

-0.34

 encantado

-0.34

 patrulla

-0.33

 embaj

-0.32

 COMPUT

-0.32

POSITIVE LOGITS

хьтан

0.69

 BoxFit

0.66

 Signalez

0.66

MLLoader

0.60

 Heisenberg

0.60

 SwitchCompat

0.57

Хьажоргаш

0.57

featureID

0.55

 ویکی‌پدی

0.55

الدراسه

0.54

Activations Density 0.008%

references to specific TV shows and their characters

TV show characters and titles

No Comments

No Known Activations