INDEX

Explanations

key terms associated with legal, emotional, and environmental contexts

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

legal terms and contracts

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

rawDesc

-0.65

 disambiguazione

-0.63

 transférez

-0.63

 zoude

-0.60

principalColumn

-0.58

SharedCtor

-0.57

tanleria

-0.57

 transfieras

-0.57

 AssemblyCulture

-0.56

 gesteld

-0.56

POSITIVE LOGITS

 démocr

0.35

urg

0.29

 Disponible

0.29

 dones

0.29

 télévis

0.28

"];

0.28

 quoi

0.27

̯

0.26

 versions

0.26

 TextEditing

0.26

Activations Density 0.073%

key terms associated with legal, emotional, and environmental contexts

legal terms and contracts

No Comments

No Known Activations