INDEX

Explanations

abbreviations, acronyms, and citations related to scientific publications

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

vig

-0.15

amma

-0.15

anga

-0.14

IES

-0.14

IfExists

-0.14

/private

-0.14

ÅĽ

-0.14

á»ĵi

-0.14

çĿĢ

-0.13

ÑģÐ¾Ð»

-0.13

POSITIVE LOGITS

oust

0.17

ABS

0.16

uet

0.15

abs

0.14

à¸Ńà¹Ģà¸¡

0.14

_TAB

0.14

çµ¶

0.14

illos

0.14

usu

0.14

à¤¾à¤¯à¤ķ

0.14

Activations Density 0.023%

abbreviations, acronyms, and citations related to scientific publications

No Comments

No Known Activations