INDEX

Explanations

specific terms and formatting indicative of legal or governmental documents

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

U.S. and United Kingdom

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

MigrationBuilder

-0.64

 saites

-0.59

styleType

-0.59

oredCriteria

-0.57

 rhestr

-0.54

 Chwiliwch

-0.54

 disambiguazione

-0.53

ImageContext

-0.51

createStatement

-0.50

RegistryLite

-0.50

POSITIVE LOGITS

Життєпис

0.37

.*")]

0.37

}`).

0.35

 nonUne

0.34

田

0.34

import

0.33

age

0.33

 mempel

0.33

IV

0.32

 пове

0.32

Activations Density 0.002%

specific terms and formatting indicative of legal or governmental documents

U.S. and United Kingdom

No Comments

No Known Activations