INDEX

Explanations

proper nouns related to significant places and individuals

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

Zandabarudo

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

 חיצוניים

-0.70

SequentialGroup

-0.68

Parcelize

-0.64

 queſta

-0.60

 cherchés

-0.60

MessageTagHelper

-0.58

iprot

-0.57

]")]

-0.57

EndInit

-0.56

ⓧ

-0.56

POSITIVE LOGITS

 hingegen

0.37

 naman

0.32

 meanwhile

0.31

̈́

0.31

 otra

0.31

 autre

0.30

 revanche

0.30

Fallback

0.29

Address

0.28

 dagegen

0.28

Activations Density 0.671%

proper nouns related to significant places and individuals

Zandabarudo

No Comments

No Known Activations