INDEX

Explanations

differently formatted or stylized sections of text, particularly symbols or unique characters

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

ä¾µ

-0.13

eoq

-0.13

ocide

-0.13

ertest

-0.12

.enterprise

-0.12

/workspace

-0.12

 à¤ĸà¤¡

-0.12

.***.***

-0.12

 INTERRUPTION

-0.11

 inadvert

-0.11

POSITIVE LOGITS

 transmissions

0.37

 transmit

0.37

 transmission

0.36

 transmitting

0.35

 Transmit

0.35

 communication

0.33

 transmitted

0.33

 Transmission

0.32

Transmission

0.32

 communicate

0.32

Activations Density 0.059%

differently formatted or stylized sections of text, particularly symbols or unique characters

No Comments

No Known Activations