INDEX

Explanations

references to academic publications and authorship details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

andin

-0.15

oker

-0.15

jej

-0.14

NCY

-0.14

Ð»Ð¸Ðº

-0.14

á»ķ

-0.14

Cum

-0.14

UGHT

-0.14

asad

-0.14

 Ð¢Ð°ÐºÐ¾Ð¶

-0.14

POSITIVE LOGITS

et

0.19

etal

0.17

eds

0.16

)null

0.16

dbl

0.15

 ÙĪÛĮÚ©ÛĮ

0.15

FTA

0.14

azzi

0.14

 Â«

0.14

eds

0.14

Activations Density 0.149%

references to academic publications and authorship details

No Comments

No Known Activations