INDEX

Explanations

terms related to television news programming and networks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

請繼續往下閱讀

-0.47

 للاسماء

-0.44

tanleria

-0.42

 ویکی‌پدیای

-0.40

cplusplus

-0.39

 يتيمه

-0.39

ourite

-0.38

httphttps

-0.37

 iNdEx

-0.37

wire

-0.37

POSITIVE LOGITS

 station

0.70

 estación

0.68

 Channel

0.65

 stations

0.63

 estaciones

0.63

station

0.57

Channel

0.57

 Station

0.56

Station

0.56

Ey

0.56

Activations Density 0.380%

terms related to television news programming and networks

No Comments

No Known Activations