INDEX

Explanations

references to news outlets and media organizations

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

Not in Any Lists

No Comments

Negative Logits

 Vintage

-0.18

Nej

-0.16

Vintage

-0.15

 vintage

-0.15

ugar

-0.15

ington

-0.14

master

-0.14

aller

-0.14

 master

-0.14

 Katz

-0.14

POSITIVE LOGITS

.scalablytyped

0.19

 Evet

0.18

anga

0.17

mey

0.15

æ¥ļ

0.15

iyel

0.15

addir

0.15

CLICK

0.14

eza

0.14

yne

0.14

Activations Density 0.035%

references to news outlets and media organizations

No Comments

No Known Activations