INDEX

Explanations

mathematical terms and relationships related to properties and proofs

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

if followed by for or they

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

jsxFileName

-0.54

 Meksiku

-0.54

PerformLayout

-0.51

AsUp

-0.49

AutoScaleMode

-0.48

 "..\..\

-0.47

:+:

-0.47

__':

-0.46

providedIn

-0.46

chartInstance

-0.45

POSITIVE LOGITS

InBackground

0.40

 means

0.37

μων

0.36

means

0.36

=*/

0.36

 dîner

0.36

Upgradeable

0.36

rieu

0.35

offsetHeight

0.35

sending

0.34

Activations Density 0.228%

mathematical terms and relationships related to properties and proofs

if followed by for or they

No Comments

No Known Activations