INDEX

Explanations

structured programs or initiatives that promote learning and collaboration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

 Hlav

-0.15

 Seks

-0.14

icz

-0.14

itler

-0.14

Å¡tÄĽ

-0.14

izens

-0.14

oha

-0.14

 Grave

-0.14

ancements

-0.13

alars

-0.13

POSITIVE LOGITS

 aims

0.21

 aimed

0.21

aim

0.19

aim

0.19

aims

0.19

 consist

0.18

 consists

0.18

 allow

0.17

 allows

0.17

æĹ

0.16

Activations Density 0.214%

structured programs or initiatives that promote learning and collaboration

No Comments

No Known Activations