INDEX

Explanations

terms and phrases that indicate comparison or distinction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

Else

-0.23

 ELSE

-0.22

 Else

-0.21

ELSE

-0.21

elif

-0.20

else

-0.20

elseif

-0.19

_else

-0.19

aft

-0.18

nev

-0.17

POSITIVE LOGITS

 others

0.33

 Others

0.25

others

0.25

Others

0.23

 other

0.22

other

0.19

0.17

 otros

0.17

 anderen

0.17

 altri

0.17

Activations Density 0.144%

terms and phrases that indicate comparison or distinction

No Comments

No Known Activations

terms and phrases that indicate comparison or distinction

No Comments

No Known Activations