INDEX

Explanations

references to competition, specifically in reality shows like "Survivor."

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

games and competitive scenarios

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

 services

-0.30

Pic

-0.28

tile

-0.28

 HTTP

-0.28

Son

-0.28

 British

-0.28

NAT

-0.28

 londres

-0.28

 republic

-0.28

fic

-0.28

POSITIVE LOGITS

 للمعارف

0.79

AnchorStyles

0.78

 kasarigan

0.70

 Geſch

0.68

WriteTagHelper

0.60

TagMode

0.59

 intptr

0.58

expliquer

0.58

RUnlock

0.58

 дописавши

0.57

Activations Density 0.004%

references to competition, specifically in reality shows like "Survivor."

games and competitive scenarios

No Comments

No Known Activations