INDEX

Explanations

code-related terms and constructs

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Alamofire

-0.46

:✨

-0.44

urlopen

-0.38

HomeAsUpEnabled

-0.38

URLException

-0.38

DialogInterface

-0.38

iprot

-0.37

StandardCharsets

-0.36

LookAndFeel

-0.35

fillStyle

-0.35

POSITIVE LOGITS

aarrggbb

0.70

parsedMessage

0.62

IsMutable

0.54

MutableLiveData

0.54

 propOrder

0.48

 Normdatei

0.48

 protoimpl

0.47

bitField

0.46

 uLocal

0.46

 ujednoznacz

0.46

Activations Density 0.004%

code-related terms and constructs

No Comments

No Known Activations