INDEX

Explanations

proper nouns, particularly names and titles related to places, characters, and products

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

proper nouns and suffixes

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

MigrationBuilder

-0.44

Vidite

-0.42

 humanidade

-0.39

 piłkarz

-0.38

 enseñar

-0.38

fasis

-0.37

afficheront

-0.37

Reprodução

-0.36

filé

-0.36

tonode

-0.36

POSITIVE LOGITS

sizeCache

0.54

 ſta

0.52

 raiſ

0.52

 purpoſe

0.52

 ſtre

0.52

 laſt

0.49

 paſſ

0.49

 ſte

0.49

Eſ

0.49

 Anſ

0.49

Activations Density 0.073%

proper nouns, particularly names and titles related to places, characters, and products

proper nouns and suffixes

No Comments

No Known Activations