INDEX

Explanations

references to groups or categories involving numbers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

Sharper

-0.53

adaptiveStyles

-0.51

-0.47

katkan

-0.47

\{\\

-0.47

 vraie

-0.46

üten

-0.46

Davidson

-0.45

 facie

-0.44

 Signore

-0.44

POSITIVE LOGITS

+#+#

1.16

 تضيفلها

0.82

ukone

0.79

bewerken

0.74

 remaining

0.70

UnusedPrivate

0.70

mybatisplus

0.68

 localVar

0.68

remaining

0.68

 aforementioned

0.67

Activations Density 0.225%

references to groups or categories involving numbers

No Comments

No Known Activations