INDEX

Explanations

terms related to skills and expertise

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

gadier

-0.49

 Palin

-0.47

 Rave

-0.45

 MessageBoxIcon

-0.44

rungsseite

-0.43

ClientSize

-0.43

 Stalker

-0.43

 pust

-0.42

endphp

-0.41

newName

-0.41

POSITIVE LOGITS

 skillful

1.02

skilled

0.94

 deft

0.93

 skilled

0.92

 skilful

0.91

 skillfully

0.90

 skil

0.87

 skill

0.85

 skills

0.83

 adept

0.82

Activations Density 0.427%

terms related to skills and expertise

No Comments

No Known Activations