INDEX

Explanations

visual elements, particularly images and diagrams

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

image and link syntax

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

AddTagHelper

-0.68

nestjs

-0.62

 ISNI

-0.61

 snippetHide

-0.60

KURZBESCHREIBUNG

-0.60

 Autorisations

-0.59

expandindo

-0.59

 AspNetCore

-0.59

transQ

-0.59

 Italijani

-0.58

POSITIVE LOGITS

 downloaded

0.44

Télécharger

0.43

 Decke

0.41

 Lombardo

0.40

 réve

0.40

這就是

0.39

 Spitze

0.38

同じく

0.38

ditto

0.37

Another

0.37

Activations Density 0.007%

visual elements, particularly images and diagrams

image and link syntax

No Comments

No Known Activations