INDEX

Explanations

programming-related terms and variables

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

omatic

-0.15

êµ°

-0.15

 Trab

-0.14

èĺŃ

-0.14

zon

-0.14

hole

-0.14

 Fill

-0.13

.tt

-0.13

 manual

-0.13

 fill

-0.13

POSITIVE LOGITS

 Fabric

0.25

Fabric

0.25

 fabric

0.24

 Channel

0.23

 channel

0.23

msp

0.22

endor

0.22

fabric

0.21

channel

0.21

 Couch

0.20

Activations Density 0.001%

programming-related terms and variables

No Comments

No Known Activations