INDEX

Explanations

references to reports and decisions related to policy and societal issues

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

words and math notation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

 Deshalb

-0.77

 كومونز

-0.66

Hence

-0.66

 stuff

-0.63

 dalamnya

-0.63

 Hence

-0.63

 سكانية

-0.62

 Kalau

-0.62

 daarvoor

-0.60

 deshalb

-0.60

POSITIVE LOGITS

 recieve

0.69

 comuniques

0.68

 recieved

0.65

0.63

 eachother

0.60

 provence

0.59

fuck

0.59

PyTuple

0.57

}:${

0.56

に於

0.55

Activations Density 0.343%

references to reports and decisions related to policy and societal issues

words and math notation

No Comments

No Known Activations