INDEX

Explanations

specific scientific terms, especially those related to neurobiology and pharmaceuticals

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

Presidential Election

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

AnimationsModule

-0.60

ioutil

-0.57

NameInMap

-0.57

 initComponents

-0.56

TypedDataSet

-0.56

exitRule

-0.55

//

-0.55

TestingModule

-0.54

SequentialGroup

-0.54

POSITIVE LOGITS

BagConstraints

0.34

 incluyendo

0.33

 precioso

0.33

yní

0.31

달

0.31

 כן

0.30

Because

0.30

 derfor

0.30

 Because

0.29

 Secara

0.29

Activations Density 0.149%

specific scientific terms, especially those related to neurobiology and pharmaceuticals

Presidential Election

No Comments

No Known Activations