INDEX

Explanations

references to URLs and login information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

bes

-0.17

phan

-0.17

gran

-0.15

ASN

-0.15

walker

-0.14

eman

-0.14

Points

-0.14

setBackground

-0.14

³

-0.14

loid

-0.14

POSITIVE LOGITS

aign

0.18

ainer

0.16

ouz

0.15

 âĢº

0.15

.ide

0.15

aille

0.14

ieu

0.14

ibo

0.14

:///

0.14

oz

0.14

Activations Density 0.004%

references to URLs and login information

No Comments

No Known Activations