INDEX

Explanations

references to studies and research reports

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

wit

-0.15

adera

-0.15

ulumi

-0.15

zept

-0.14

_UNS

-0.14

stoff

-0.14

ãĥ

-0.14

 misd

-0.13

_begin

-0.13

adders

-0.13

POSITIVE LOGITS

 released

0.36

commission

0.34

 commissioned

0.31

 commission

0.30

released

0.30

 published

0.27

 Released

0.27

Released

0.26

 Commission

0.26

co

0.26

Activations Density 0.071%

references to studies and research reports

No Comments

No Known Activations