INDEX

Explanations

technical layout and design terms in coding and user interface contexts

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

false discovery rate, Benjamini-Hochberg

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

esModule

-0.57

sendRedirect

-0.56

antd

-0.51

typeorm

-0.51

himovic

-0.48

ագրություններ

-0.47

 ModelAndView

-0.46

GetMapping

-0.46

withIdentifier

-0.45

lamydia

-0.44

POSITIVE LOGITS

ChildScrollView

0.78

wrap

0.64

 wrap

0.49

WRAP

0.48

Wrap

0.44

elemField

0.42

 Wrap

0.41

uação

0.40

 suivez

0.40

IsMutable

0.37

Activations Density 0.001%

technical layout and design terms in coding and user interface contexts

false discovery rate, Benjamini-Hochberg

No Comments

No Known Activations