INDEX

Explanations

the presence of a specific term or name repeatedly

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

reu

-0.17

Rei

-0.16

 Ð¿ÐµÑĢÐµÑĢ

-0.15

ilar

-0.14

alfa

-0.14

lando

-0.14

ignite

-0.14

à¹ĥà¸Ī

-0.14

iciar

-0.13

Å¡ov

-0.13

POSITIVE LOGITS

ively

0.29

utilus

0.28

ive

0.27

ught

0.24

omi

0.24

ughty

0.23

UGHT

0.21

IVE

0.20

oto

0.19

issance

0.19

Activations Density 0.017%

the presence of a specific term or name repeatedly

No Comments

No Known Activations