INDEX

Explanations

elements related to membership or organization references

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

theless

-0.30

plier

-0.29

thing

-0.26

Ø©

-0.25

ember

-0.24

à¸ģà¸²à¸£

-0.22

cluding

-0.21

Øª

-0.21

DAQ

-0.20

uary

-0.20

POSITIVE LOGITS

uards

0.20

íļĮìĿĺ

0.17

ards

0.15

ungs

0.15

 Wolff

0.15

ÅĽnie

0.15

bbb

0.14

-wsj

0.14

irst

0.14

Falsy

0.14

Activations Density 0.452%

elements related to membership or organization references

No Comments

No Known Activations