INDEX

Explanations

references to mountains and elevated geographical features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

#--

-0.36

darwin

-0.33

testid

-0.32

Qa

-0.32

Nickname

-0.32

 Jeffer

-0.32

-0.31

 Sears

-0.31

 اص

-0.31

Tottenham

-0.30

POSITIVE LOGITS

 Mountain

0.81

 River

0.77

Mountain

0.77

 Island

0.75

 MOUNTAIN

0.71

Portail

0.71

River

0.71

 mountain

0.71

 river

0.70

 island

0.69

Activations Density 0.254%

references to mountains and elevated geographical features

No Comments

No Known Activations