INDEX

Explanations

syntax and structure related to Python code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

RTE

-0.14

LIBINT

-0.14

couz

-0.13

Éµ

-0.13

atin

-0.12

ëŀĢ

-0.12

 nackte

-0.12

 Strauss

-0.12

parseFloat

-0.12

=fopen

-0.12

POSITIVE LOGITS

def

0.40

 print

0.34

def

0.34

print

0.32

#print

0.31

 import

0.30

>>>

0.27

-def

0.26

#create

0.25

Activations Density 0.047%

syntax and structure related to Python code

No Comments

No Known Activations