INDEX

Explanations

patterns of whitespace and indentation in formatted code or text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

pon

-0.20

arra

-0.19

iyeti

-0.15

.mk

-0.15

 Edison

-0.15

ETY

-0.14

istrovstvÃŃ

-0.14

ç´

-0.14

æ¹¾

-0.14

uset

-0.14

POSITIVE LOGITS

ortho

0.15

-urlencoded

0.14

-css

0.14

asse

0.14

-relative

0.14

park

0.13

CLK

0.13

lle

0.13

 lodging

0.13

gt

0.13

Activations Density 0.030%

patterns of whitespace and indentation in formatted code or text

No Comments

No Known Activations