INDEX

Explanations

ethical considerations and declarations in research contexts

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

weigh in favor of

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

 autorytatywna

-0.98

+#+#

-0.91



-0.87

Autoritní

-0.87

AddTagHelper

-0.82

principalTable

-0.82

 wireType

-0.81

Попис

-0.80

RTSC

-0.77

 صوتيه

-0.77

POSITIVE LOGITS

 soulign

0.35

 confiança

0.33

 undermined

0.31

exécution

0.30

 Aplica

0.29

 aplican

0.29

 pembunuhan

0.28

 aplicação

0.27

 parecer

0.27

 seleção

0.27

Activations Density 0.037%

ethical considerations and declarations in research contexts

weigh in favor of

No Comments

No Known Activations