INDEX

Explanations

instances of code or functions related to data structures and processing in programming contexts

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

code changes and comments

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

 بيها

-0.61

ExtendWith

-0.55

ArrowToggle

-0.55

$_['

-0.53

RectangleBorder

-0.52

 AspNetCore

-0.51

kusen

-0.50

 సౌకర్య

-0.50

expandindo

-0.50

AddTagHelper

-0.49

POSITIVE LOGITS

 Vergrößern

0.42

0.39

Dire

0.39

 surla

0.35

 heran

0.35

帖最后由

0.35

Plus

0.33

artig

0.33

 macros

0.33

?!

0.32

Activations Density 0.011%

instances of code or functions related to data structures and processing in programming contexts

code changes and comments

No Comments

No Known Activations