INDEX

Explanations

programming-related keywords and code structure

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

 aDecoder

-0.47

__(/*!

-0.46

帖最后由

-0.45

MLLoader

-0.44

nmgp

-0.41

 Barbier

-0.41

<?

-0.41

UrlResolution

-0.40

 listo

-0.39

InjectMocks

-0.39

POSITIVE LOGITS

NSInteger

0.85

BOOL

0.46

BNB

0.40

AndEndTag

0.39

`}`,

0.38

 BOOL

0.37

وفاته

0.37

iffen

0.37

iences

0.36

 Sentin

0.36

Activations Density 0.000%

programming-related keywords and code structure

No Comments

No Known Activations