INDEX

Explanations

expressions related to consequences and accountability in a judicial context

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

setting and characters

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

 فريبيس

-0.92

 Lithuan

-0.66

 Monfieur

-0.65

 Majefty

-0.65

 transfieras

-0.65

 Komunikasi

-0.60

 PeEnEo

-0.59

 Kasih

-0.58

getMenuInflater

-0.58

 Anſ

-0.58

POSITIVE LOGITS

 hindurch

0.33

 BrowserModule

0.33

LinkId

0.30

 town

0.28

 Wheel

0.28

 pelatihan

0.28

 générales

0.27

tam

0.26

 menangkap

0.26

fällig

0.26

Activations Density 0.101%

expressions related to consequences and accountability in a judicial context

setting and characters

No Comments

No Known Activations