INDEX

Explanations

key terms related to categorization and classification

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

not automatically called, convergence

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

 Савезне

-0.57

tagHelperRunner

-0.50

labus

-0.49

 Biôgrafia

-0.48

-------

-0.47

DropTable

-0.47

extAlignment

-0.47

Див

-0.46

richTextPanel

-0.46

bkz

-0.46

POSITIVE LOGITS

Handlung

0.35

 insufficient

0.32

 fallen

0.32

setCellStyle

0.31

 недоста

0.31

Background

0.31

lui

0.31

 THOMPSON

0.31

phy

0.31

 kors

0.31

Activations Density 0.214%

key terms related to categorization and classification

not automatically called, convergence

No Comments

No Known Activations