INDEX

Explanations

terms related to health-related behaviors and conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

lò

-0.38

 penger

-0.38

internalType

-0.38

usammen

-0.36

ỡng

-0.35

 işle

-0.35

mybatisplus

-0.35

 tekem

-0.34

 melang

-0.32

.*")]

-0.32

POSITIVE LOGITS

 houſe

0.66

 epidemiological

0.65

 <<<<<<<<<<<<<<

0.60

 nationally

0.59

 study

0.59

 ſeveral

0.58

kháu

0.57

 epidemiology

0.57

 epidemio

0.55

 ſtate

0.54

Activations Density 2.723%

terms related to health-related behaviors and conditions

No Comments

No Known Activations