INDEX

Explanations

phrases indicating action or states of being, particularly concerning desire and conflict within relationships

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 kajian

-0.32

 partai

-0.30

berdayakan

-0.30

 bēr

-0.29

 haberse

-0.29

 kvinna

-0.29

 sucedido

-0.28

 kości

-0.28

 tradisi

-0.28

by

-0.28

POSITIVE LOGITS

 queſta

0.93

 Мексичка

0.92

SourceChecksum

0.87

InitVars

0.87

 otomatig

0.87

 gynhyrchwyd

0.86

<unused79>

0.85

<unused41>

0.85

<unused8>

0.85

[@BOS@]

0.85

Activations Density 0.815%

phrases indicating action or states of being, particularly concerning desire and conflict within relationships

No Known Activations