INDEX

Explanations

terms related to architecture

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 pleaſure

-1.08

Efq

-0.99

 whoſe

-0.94

 Theſe

-0.92

 Beſ

-0.92

 ſever

-0.91

 ſeveral

-0.91

 myſelf

-0.91

 Anſ

-0.88

 leſs

-0.87

POSITIVE LOGITS

 arch

1.23

 Arch

1.18

Ar

1.14

 Arches

1.11

Ar

1.07

Arch

1.03

ar

1.02

 ARCH

1.00

arc

0.99

 arches

0.98

Activations Density 0.092%

terms related to architecture

No Comments

No Known Activations