INDEX

Explanations

changes in rates or measurements related to performance and efficiency

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

contentLoaded

-0.65

enumi

-0.60

 Infórmanos

-0.51

 الاطلاع

-0.50

 transfieras

-0.50

InSection

-0.49

deelte

-0.49

 defStyle

-0.49

setVerticalGroup

-0.49

ushan

-0.48

POSITIVE LOGITS

 speed

2.56

 speeds

2.25

speed

2.19

 pace

2.13

 Speed

2.06

Speed

2.04

 rate

2.02

 faster

2.00

 vitesse

2.00

 SPEED

1.95

Activations Density 0.843%

changes in rates or measurements related to performance and efficiency

No Comments

No Known Activations