INDEX

Explanations

references to coffee and coffee-related establishments

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

coffee and barista

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eraard

-0.39

blauch

-0.39

 carnes

-0.38

ReusableCell

-0.38

 ràng

-0.38

 diverti

-0.37

 divertimento

-0.37

 Italijanski

-0.37

荤

-0.37

setVerticalGroup

-0.36

POSITIVE LOGITS

 coffee

0.90

Coffee

0.85

coffee

0.83

☕️

0.77

 Coffee

0.77

 espresso

0.77

 caffeine

0.77

 COFFEE

0.75

 barista

0.72

 koffie

0.71

Activations Density 0.025%

references to coffee and coffee-related establishments

coffee and barista

No Comments

No Known Activations