INDEX

Explanations

phrases related to predictability and expectation

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

predictability and regularity

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 Consulte

-0.38

 indépendant

-0.38

 exceptionnel

-0.36

 abstrait

-0.35

 quedado

-0.35

 realizarse

-0.35

 Geschichts

-0.35

 considération

-0.34

Important

-0.34

 producteur

-0.34

POSITIVE LOGITS

 predictability

0.59

 predic

0.58

 predictable

0.58

 Predic

0.54

WebElementEntity

0.53

predic

0.53

 prediction

0.52

findpost

0.52

 MainAxisSize

0.51

цездатний

0.51

Activations Density 0.143%

phrases related to predictability and expectation

predictability and regularity

No Comments

No Known Activations