INDEX

Explanations

punctuation marks and mathematical notations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

neck

-0.16

eron

-0.14

rex

-0.14

 Press

-0.14

Asked

-0.14

REM

-0.13

Rus

-0.13

 fare

-0.13

rek

-0.13

opo

-0.13

POSITIVE LOGITS

tainment

0.16

end

0.15

ÏĦÎµÏģ

0.15

oler

0.15

à¸Ĺà¸Ńà¸ĩ

0.15

","\

0.15

æ´

0.14

CACHE

0.14

odium

0.14

Activations Density 0.053%

punctuation marks and mathematical notations

No Comments

No Known Activations