INDEX

Explanations

elements that express positivity and supportiveness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 indeed

-0.24

 only

-0.21

 atleast

-0.20

only

-0.20

 Ð¸Ð¼ÐµÐ½Ð½Ð¾

-0.19

 quite

-0.19

 both

-0.18

 neither

-0.18

 BOTH

-0.18

 ONLY

-0.18

POSITIVE LOGITS

 plain

0.23

 thÃ´i

0.23

ifiable

0.22

ifi

0.20

ifying

0.20

IFI

0.19

æĻ®éĢļ

0.18

plain

0.18

 Plain

0.18

vailability

0.17

Activations Density 0.167%

elements that express positivity and supportiveness

No Comments

No Known Activations