INDEX

Explanations

functionalities related to file handling or writing processes in a programming context

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

code tests and structures

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

̓

-0.43

TypedDataSet

-0.42

 EconPapers

-0.41

 Wicidata

-0.39

exitRule

-0.39

 Vietnamese

-0.38

 Armenian

-0.38

 ModelExpression

-0.38

awtextra

-0.36

estacks

-0.36

POSITIVE LOGITS

WaitGroup

0.92

AddTagHelper

0.48

 propOrder

0.46

Sklici

0.45

 disambiguazione

0.42

UserScript

0.42

цездатний

0.40

glTexCoord

0.39

ทธิ

0.39

 negócios

0.39

Activations Density 0.023%

functionalities related to file handling or writing processes in a programming context

code tests and structures

No Comments

No Known Activations