INDEX

Explanations

phrases referring to conditional or causative statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

Loo

-0.47

 anſ

-0.46

⭑

-0.46

Loo

-0.45

 isOpen

-0.45

 vrijwilli

-0.44

 fisico

-0.44

 Loop

-0.43

 loop

-0.43

 raiſ

-0.43

POSITIVE LOGITS

that

1.03

 that

1.01

THAT

1.00

That

0.99

 That

0.96

 THAT

0.94

 bahwa

0.83

 bahawa

0.81

que

0.76

 rằng

0.75

Activations Density 0.632%

phrases referring to conditional or causative statements

No Comments

No Known Activations