INDEX

Explanations

references to projects or initiatives

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

project and its context

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

 ancianos

-0.55

 comerciantes

-0.51

 berapa

-0.50

 Савезне

-0.49

 defStyleAttr

-0.49

authenticated

-0.48

 bezpeč

-0.48

Cuándo

-0.48

 trasparente

-0.47

 soixante

-0.47

POSITIVE LOGITS

 Project

1.17

Project

1.12

 project

1.10

project

1.05

 projec

1.03

 PROJECT

0.98

PROJECT

0.91

ProjectId

0.91

 Projects

0.89

getProject

0.89

Activations Density 0.015%

references to projects or initiatives

project and its context

No Comments

No Known Activations