INDEX

Explanations

references to particular seasons or episodes within a series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

Autowired

-0.58

 Karls

-0.52

Autres

-0.52

utuhkan

-0.51

 Karlsruhe

-0.49

WebServlet

-0.48

 Swartz

-0.48

 Kowalski

-0.47

 white

-0.47

 Smart

-0.47

POSITIVE LOGITS

 season

0.92

season

0.92

 Season

0.89

 SEASON

0.88

Season

0.86

seasons

0.83

SEASON

0.79

 seasons

0.75

 seasonally

0.74

 Seasons

0.73

Activations Density 0.167%

references to particular seasons or episodes within a series

No Comments

No Known Activations