INDEX

Explanations

academic qualifications and affiliations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

Å¥

-0.19

 Zones

-0.18

utter

-0.16

andard

-0.16

 slee

-0.15

 Millet

-0.14

è¨Ģãģ£ãģŁ

-0.14

 bure

-0.14

chner

-0.14

adian

-0.13

POSITIVE LOGITS

Fac

0.26

 Babe

0.22

 Faculty

0.21

 polite

0.21

Fac

0.20

 Techn

0.20

 ÑĦÐ°Ðº

0.20

Jag

0.20

 Univers

0.19

 Technical

0.19

Activations Density 0.077%

academic qualifications and affiliations

No Comments

No Known Activations