INDEX

Explanations

references to historical or cultural heritage listings

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

listed or registered items

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

 global

-0.38

 rush

-0.36

WithMany

-0.35

 char

-0.35

 concept

-0.34

 cloud

-0.34

 bottom

-0.32

 self

-0.31

 StringField

-0.31

mat

-0.30

POSITIVE LOGITS

 Wicidata

0.73

 Waſſer

0.72

 EconPapers

0.71

 geſch

0.64

 Mérimée

0.63

 disambiguazione

0.62

<unused41>

0.61

<unused51>

0.61

<unused14>

0.61

[@BOS@]

0.61

Activations Density 0.009%

references to historical or cultural heritage listings

listed or registered items

No Comments

No Known Activations