INDEX

Explanations

references to "village" and its variations

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

village names

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

-0.51

on

-0.48

should

-0.47

-0.46

He

-0.46

$\

-0.45

 McKinnon

-0.45

-0.44

(\

-0.43

no

-0.43

POSITIVE LOGITS

 village

1.33

 Village

1.29

village

1.23

Village

1.22

 villages

1.14

 VILLAGE

1.13

 Villages

1.08

 villaggio

0.94

 villager

0.94

 villagers

0.85

Activations Density 0.005%

references to "village" and its variations

village names

No Comments

No Known Activations