INDEX

Explanations

academic references and citations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

benh

-0.19

utral

-0.17

arcy

-0.15

itus

-0.15

ourn

-0.15

lint

-0.15

ITO

-0.14

é¾į

-0.14

ihan

-0.14

ibling

-0.14

POSITIVE LOGITS

_UNS

0.16

oba

0.16

ocos

0.14

aris

0.14

Khu

0.14

Ã²a

0.13

ju

0.13

olit

0.13

/releases

0.13

stell

0.13

Activations Density 0.096%

academic references and citations

No Comments

No Known Activations