INDEX

Explanations

sequences of numeric values

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

resolutions and dimensions

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

VersionUID

-0.47

PARATUS

-0.47

AsUp

-0.45

ChildIndex

-0.45

 للمعارف

-0.44

WriteTagHelper

-0.42

 heapq

-0.41

RegressionTest

-0.41

LookAnd

-0.41

 informée

-0.40

POSITIVE LOGITS

 AssemblyCulture

0.42

ChildScrollView

0.41

 ویکی‌پدی

0.40

NVIC

0.40

XtraGrid

0.39

Sklici

0.39

udios

0.39

ofte

0.38

 synthetic

0.37

setVerticalGroup

0.37

Activations Density 0.029%

sequences of numeric values

resolutions and dimensions

No Comments

No Known Activations