INDEX

Explanations

instructions, warnings, and details about operational performance or issues with devices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

à¸Ńà¸Ķ

-0.15

Suc

-0.15

akin

-0.14

 confess

-0.14

amat

-0.14

 patch

-0.14

dil

-0.14

 patches

-0.14

edImage

-0.14

 orders

-0.13

POSITIVE LOGITS

Refer

0.17

 Refer

0.16

refer

0.15

 groove

0.15

ambre

0.15

pong

0.15

skyt

0.14

ante

0.14

cket

0.14

akk

0.14

Activations Density 0.018%

instructions, warnings, and details about operational performance or issues with devices

No Comments

No Known Activations