INDEX

Explanations

references to capital and capital expenditures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

<bos>

-0.42

 actionPerformed

-0.36

Бележки

-0.36

assertIn

-0.36

lihood

-0.35

ness

-0.35

 effectué

-0.35

Източници

-0.35

إنه

-0.35

czaj

-0.34

POSITIVE LOGITS

Capital

0.70

 Capital

0.68

 capital

0.68

 capitale

0.64

 CAPITAL

0.62

capital

0.60

 capitol

0.56

 raiſ

0.55

CAPITAL

0.55

 Capitals

0.54

Activations Density 0.175%

references to capital and capital expenditures

No Comments

No Known Activations

references to capital and capital expenditures

No Comments

No Known Activations