INDEX

Explanations

phrases indicating a caution or important information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

+#+#

-0.53

 houſe

-0.52

 <<<<<<<<<<<<<<

-0.51

UnitTesting

-0.47

 Houſe

-0.47

 smtplib

-0.44

Err

-0.42

 nephe

-0.42

 UnityEditor

-0.41

erm

-0.41

POSITIVE LOGITS

Notice

0.72

 Notice

0.69

 notice

0.67



0.64

notice

0.60

注意

0.59

 注意

0.58

 Note

0.57

Note

0.57

 note

0.56

Activations Density 0.344%

phrases indicating a caution or important information

No Comments

No Known Activations