INDEX

Explanations

numerical references and data points, particularly relating to years in historical contexts

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

ibbon

-0.15

chten

-0.14

lot

-0.14

ooter

-0.14

ÑĥÑģÑĤ

-0.14

Ð¾ÑģÑĤÑĥÐ¿

-0.14

lot

-0.13

esture

-0.13

ooting

-0.13

smouth

-0.13

POSITIVE LOGITS

Rum

0.18

IDES

0.15

dsl

0.15

 Hughes

0.15

QualifiedName

0.14

 è¯

0.14

 Cuomo

0.14

SCORE

0.14

 Bomb

0.14

des

0.14

Activations Density 0.003%

numerical references and data points, particularly relating to years in historical contexts

No Comments

No Known Activations