INDEX

Explanations

alphanumeric sequences and identifiers, likely related to URLs or technical codes

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

lac

-0.16

ileo

-0.15

ihat

-0.15

ÐµÐ»Ñİ

-0.15

aca

-0.15

Ã©is

-0.14

antis

-0.14

æĬŀ

-0.14

emos

-0.14

AndWait

-0.14

POSITIVE LOGITS

_INLINE

0.16

0.15

\Catalog

0.14

éĩı

0.14

Pep

0.13

-edit

0.13

 Archives

0.13

directive

0.13

emics

0.13

IMENT

0.13

Activations Density 0.065%

alphanumeric sequences and identifiers, likely related to URLs or technical codes

No Comments

No Known Activations