INDEX

Explanations

references to meals and dining experiences

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

 agência

-0.47

 regalías

-0.47

 digitais

-0.46

 judiciales

-0.45

 McKenzie

-0.44

 vouloir

-0.44

 juges

-0.42

 sanitaires

-0.42

zeß

-0.41

 behaupten

-0.41

POSITIVE LOGITS

Dinner

0.98

 Dinner

0.97

 dinner

0.95

dinner

0.92

 lunch

0.89

 Lunch

0.89

 DINNER

0.86

Lunch

0.85

lunch

0.82

 breakfast

0.81

Activations Density 0.047%

references to meals and dining experiences

No Comments

No Known Activations