INDEX

Explanations

terms and phrases related to transformation and change in context

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

actions and transformations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

ness

-0.37

body

-0.36

ism

-0.33

 body

-0.32

iness

-0.32

math

-0.32

wall

-0.32

WEBPACK

-0.31

_{\

-0.30

جوايز

-0.30

POSITIVE LOGITS

 ujednoznacz

0.85

 يتيمه

0.84

IntoConstraints

0.78

DIPSETTING

0.72

 <>",

0.71

 Italijani

0.70

 disambiguazione

0.69

 ویکی‌پدیا

0.69

UserScript

0.69

 المعيارى

0.69

Activations Density 0.110%

terms and phrases related to transformation and change in context

actions and transformations

No Comments

No Known Activations