INDEX

Explanations

elements related to problem-solving or inquiry

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

in languages other than English

np_acts-logits-general · gemini-2.0-flash

scientific and numerical data

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

código

-0.68

 Teléfono

-0.64

 dacă

-0.64

să

-0.64

 cuánt

-0.64

 Māori

-0.63

În

-0.63

 Dacă

-0.63

portál

-0.63

 élite

-0.63

POSITIVE LOGITS

 wasnt

0.61

 isnt

0.60

Ã

0.58

__

0.58

 doesnt

0.58

 Joao

0.57

�

0.56

 Francois

0.56

 proposito

0.56

 didnt

0.56

Activations Density 0.352%

elements related to problem-solving or inquiry

in languages other than English

scientific and numerical data

No Comments

No Known Activations