INDEX

Explanations

terms related to implementation details and technical specifications

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

implementation and alternatives

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

ItemBackground

-0.48

 onView

-0.45

 brillo

-0.45

DropTable

-0.44

 оригіналу

-0.44

 AttributeSet

-0.43

 незавершена

-0.42

neck

-0.42

쩌

-0.41

 unknownFields

-0.41

POSITIVE LOGITS

 ExecuteAsync

0.46

 године

0.46

WithIOException

0.43

 utafitiHapana

0.41

 transfieras

0.36

 invokingState

0.35

chauung

0.34

matchCondition

0.34

thenReturn

0.34

SourceChecksum

0.33

Activations Density 0.063%

terms related to implementation details and technical specifications

implementation and alternatives

No Comments

No Known Activations