INDEX

Explanations

phrases or tokens related to data structure or reference formats

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

figures, tables, and notes

np_acts-logits-general · gemini-2.0-flash

references in brackets

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

Geografía

-0.63

stateProvider

-0.52

↵

-0.51

 useCallback

-0.48

...

-0.48

********/

-0.48

-0.47

…

-0.47

Основ

-0.47

tre

-0.47

POSITIVE LOGITS

脚注の使い方

0.85

 Савезне

0.83

expandindo

0.73

ſelves

0.73

StoryboardSegue

0.72

 utafitiHapana

0.69

 createState

0.68

 Majefty

0.66

مصادر

0.65

 purpoſe

0.65

Activations Density 0.094%

phrases or tokens related to data structure or reference formats

figures, tables, and notes

references in brackets

No Comments

No Known Activations