INDEX

Explanations

references to geographic locations or specific places

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

Pab

-0.51

Tro

-0.51

Vis

-0.50

Gro

-0.49

Flu

-0.49

СТВ

-0.48

Sab

-0.48

Pro

-0.47

Vic

-0.46

 vVar

-0.46

POSITIVE LOGITS

expandindo

0.68

 laikā

0.61

 medži

0.60

 FetchType

0.60

曖昧さ回避

0.59

 juos

0.58

 izdel

0.57

 civilización

0.57

 Italij

0.56

 cuchillo

0.56

Activations Density 0.039%

references to geographic locations or specific places

No Comments

No Known Activations