INDEX

Explanations

comparative phrases emphasizing "more than."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 Gass

-0.76

 Jäger

-0.69

bufio

-0.66

Przypisy

-0.66

Busch

-0.66

 gills

-0.61

췄

-0.60

 koper

-0.60

appropri

-0.59

isser

-0.59

POSITIVE LOGITS

 THAN

1.02

 than

0.90

 Than

0.90

iastes

0.84

Than

0.81

CloseOperation

0.80

 dozen

0.79

THAN

0.78

ymce

0.78

demies

0.78

Activations Density 0.075%

comparative phrases emphasizing "more than."

No Comments

No Known Activations