INDEX

Explanations

technical terms related to data processing and interpolation methods

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

interpolation parameter, flex item, etching target, transform coefficients

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

ClientSize

-0.48

DockStyle

-0.46

 wuß

-0.45

 człowieka

-0.41

printStackTrace

-0.41

 země

-0.41

的她

-0.41

醐

-0.40

 coû

-0.40

GOTREF

-0.40

POSITIVE LOGITS

 للمعارف

0.57

 accounts

0.44

zan

0.43

bindingNavigator

0.43

 chamber

0.43

 accession

0.42

dafx

0.42

 cross

0.42

hbase

0.41

Popis

0.41

Activations Density 0.252%

technical terms related to data processing and interpolation methods

interpolation parameter, flex item, etching target, transform coefficients

No Comments

No Known Activations