INDEX

Explanations

instances related to personal identities and their implications

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

Japanese nouns followed by particles

np_acts-logits-general · gemini-2.0-flash

situation, human, event, trace

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

хьтан

-0.69

:✨

-0.68

WriteBarrier

-0.59

sidemargin

-0.55

nezeu

-0.54

onViewCreated

-0.54

adaptiveStyles

-0.53

Chham

-0.52

цездатний

-0.52

rrggbb

-0.50

POSITIVE LOGITS

は

0.77

が

0.66

에는

0.64

には

0.64

 entanto

0.63

은

0.63

 때문에

0.62

는

0.61

はその

0.60

はは

0.59

Activations Density 0.052%

instances related to personal identities and their implications

Japanese nouns followed by particles

situation, human, event, trace

No Comments

No Known Activations