INDEX

Explanations

processes and methods related to technical troubleshooting and debugging

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

kers

-0.15

itas

-0.15

Sed

-0.15

iversal

-0.15

icas

-0.14

asha

-0.14

gra

-0.14

.EventQueue

-0.14

affles

-0.14

itzer

-0.14

POSITIVE LOGITS

lyph

0.17

alara

0.15

raquo

0.14

Ø§ØªØ±

0.14

CEEDED

0.14

ramid

0.14

 Ð²ÑĭÐ²Ð¾Ð´

0.14

icode

0.14

raph

0.14

stad

0.14

Activations Density 0.127%

processes and methods related to technical troubleshooting and debugging

No Known Activations