INDEX

Explanations

phrases related to programming concepts and logic

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

Followed by verbs in other languages

np_acts-logits-general · gemini-2.0-flash

programming variables and actions

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

 weren

-0.82

 murieron

-0.82

 pudieran

-0.81

 abbiano

-0.80

tagHelperRunner

-0.78

InvalidProtocol

-0.77

 devront

-0.76

 pouvaient

-0.76

 aient

-0.74

 puedas

-0.73

POSITIVE LOGITS

se

0.74

le

0.66

Personensuche

0.65

 assume

0.56

ha

0.56

eventbus

0.53

re

0.52

me

0.46

no

0.45

exp

0.45

Activations Density 0.060%

phrases related to programming concepts and logic

Followed by verbs in other languages

programming variables and actions

No Comments

No Known Activations