INDEX

Explanations

questions or phrases that emphasize the word "what."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

es

-0.60

 Moors

-0.58

☆☆

-0.58

ください

-0.57

 brook

-0.57

Brooks

-0.57

 Astral

-0.57

 BROOK

-0.56

ారు

-0.56

lanır

-0.56

POSITIVE LOGITS

 what

1.48

 WHAT

1.47

WHAT

1.42

 What

1.35

what

1.34

What

1.32

Hvad

0.87

Hva

0.87

ndham

0.86

 happened

0.86

Activations Density 0.127%

questions or phrases that emphasize the word "what."

No Comments

No Known Activations