INDEX

Explanations

terms related to hierarchy or classification systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.39.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

 gekomen

-0.35

 geweest

-0.34

 precum

-0.29

 mijne

-0.29

 entera

-0.27

 both

-0.26

 afhankelijk

-0.26

 nimic

-0.26

 hayan

-0.25

 verschill

-0.25

POSITIVE LOGITS

 queſta

1.13

parsedMessage

1.08

 nahilalakip

1.00

niſſe

0.98

ſammen

0.97

ſelben

0.96

<unused52>

0.94

 ſind

0.94

 betweenstory

0.94

<unused14>

0.93

Activations Density 3.935%

terms related to hierarchy or classification systems

No Comments

No Known Activations