INDEX

Explanations

attributions and editorial contributions in academic writing

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

academic editing by

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

 CreateTagHelper

-0.71

 ſche

-0.58

 المعيارى

-0.54

 itſelf

-0.52

 pleaſure

-0.50

 ſtand

-0.50

 ſch

-0.49

يميديا

-0.49

 متعلقه

-0.48

 becauſe

-0.47

POSITIVE LOGITS

 editing

0.56

 Editing

0.54

 edited

0.51

 Editor

0.50

Editor

0.50

Editing

0.49

 edit

0.47

DeleteBehavior

0.47

 editor

0.46

 Editors

0.45

Activations Density 0.005%

attributions and editorial contributions in academic writing

academic editing by

No Comments

No Known Activations