INDEX

Explanations

references to stages and responses in structured data or classification systems

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

specific list items

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

SharedCtor

-0.49

thâu

-0.48

/*

-0.47

BufferException

-0.46

errHandler

-0.44

NameInMap

-0.43

Дереккөздер

-0.42

 GenerationType

-0.41

endphp

-0.41

 surla

-0.38

POSITIVE LOGITS

principalTable

0.66

rungsseite

0.53

ParallelGroup

0.50

 lastly

0.46

CppMethod

0.44

HomeAsUpEnabled

0.42

最后

0.41

imageshack

0.41

 finally

0.40

-------------</

0.40

Activations Density 0.030%

references to stages and responses in structured data or classification systems

specific list items

No Comments

No Known Activations