INDEX

Explanations

formatting and structure indicators in code snippets

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

programming language keywords and constructs

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

✎

-0.47

 CanadaChoose

-0.47

jell

-0.40

actory

-0.40

 ddelweddau

-0.39

oplayer

-0.39

findpost

-0.38

طوب

-0.38

privateKey

-0.38

DECREF

-0.38

POSITIVE LOGITS

Personendaten

0.56

WithMany

0.46

 Савезне

0.44

NamedQueries

0.43

 Roskov

0.40

ModelSerializer

0.39

Unmarshaller

0.39

__':

0.37

原始内容

0.36

 findOne

0.36

Activations Density 0.001%

formatting and structure indicators in code snippets

programming language keywords and constructs

No Comments

No Known Activations