INDEX

Explanations

references to buildings

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

describes buildings

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 kasarigan

-0.66

 mergeFrom

-0.59

TagMode

-0.59

 Geiſt

-0.55

iſchen

-0.54

wpi

-0.54

 Schrö

-0.53

</tfoot>

-0.53

CppMethod

-0.52

wantErr

-0.52

POSITIVE LOGITS

 building

1.35

 Building

1.34

Building

1.30

 buildings

1.28

 Buildings

1.27

 edificio

1.22

building

1.20

Buildings

1.18

 BUILDING

1.16

buildings

1.13

Activations Density 0.007%

references to buildings

describes buildings

No Comments

No Known Activations