INDEX

Explanations

references to scholarly or academic articles and publications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 Walters

-0.17

appropri

-0.14

acer

-0.14

 Ã¶den

-0.13

 Schro

-0.13

åħ¹

-0.12

 Loose

-0.12

Ø

-0.12

 porad

-0.12

/browse

-0.12

POSITIVE LOGITS

0.16

serie

0.15

.qq

0.15

hem

0.15

apo

0.15

Î¸Îµ

0.15

 âĹĦ

0.14

 Antar

0.14

Activations Density 0.045%

references to scholarly or academic articles and publications

No Comments

No Known Activations