INDEX

Explanations

programming-related keywords and syntax

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

 patriot

-0.18

 Pearce

-0.18

 Polar

-0.17

 popover

-0.17

Planet

-0.16

ignon

-0.16

 Patterson

-0.16

/pm

-0.16

Pont

-0.15

 patriotism

-0.15

POSITIVE LOGITS

-p

0.33

ãĥĳ

0.26

-P

0.23

 Ð¿

0.23

 cubic

0.21

pu

0.21

pul

0.20

 cÃ´ng

0.19

 ãĥĳ

0.19

oub

0.18

Activations Density 0.083%

programming-related keywords and syntax

No Comments

No Known Activations