INDEX

Explanations

structured data or keywords relating to organizational and systematic components

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

.TODO

-0.15

yal

-0.15

Ã¡st

-0.15

 Cook

-0.15

Twe

-0.14

 Heights

-0.14

irie

-0.13

hell

-0.13

Border

-0.13

okie

-0.13

POSITIVE LOGITS

oud

0.16

oggle

0.15

isque

0.14

inton

0.14

cta

0.14

PTY

0.14

cyk

0.13

leh

0.13

 cdecl

0.13

 Burk

0.13

Activations Density 0.080%

structured data or keywords relating to organizational and systematic components

No Comments

No Known Activations