INDEX

Explanations

variations of the word "rem" related to reminders or references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

 compro

-0.43

Fernseh

-0.43

 Waugh

-0.38

 traveled

-0.37

helin

-0.37

горя

-0.37

onedDateTime

-0.36

tokyo

-0.36

 MatDialog

-0.36

 judo

-0.35

POSITIVE LOGITS

Rem

0.86

Rem

0.77

rem

0.73

REM

0.72

rem

0.69

REM

0.66

ConstraintMaker

0.60

 Remy

0.59

enumi

0.58

 Rema

0.57

Activations Density 0.178%

variations of the word "rem" related to reminders or references

No Comments

No Known Activations