INDEX

Explanations

references to concealment or hiding

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

DockStyle

-0.54

ſelf

-0.51

URLConnection

-0.50

CharStream

-0.46

exclu

-0.45

 deck

-0.43

inv

-0.43

 deſ

-0.43

 deſt

-0.42

 Lenz

-0.42

POSITIVE LOGITS

 Hide

0.83

 hide

0.83

 incident

0.82

Hide

0.80

 Incident

0.74

incident

0.70

thing

0.69

hide

0.63

 thing

0.63

 hiding

0.63

Activations Density 0.089%

references to concealment or hiding

No Comments

No Known Activations