INDEX

Explanations

terms related to resource allocation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

Å¡i

-0.16

ackle

-0.15

ux

-0.15

acob

-0.15

uzz

-0.15

civ

-0.14

isman

-0.14

 Declarations

-0.14

 Medieval

-0.14

ë¬´

-0.13

POSITIVE LOGITS

agrid

0.17

oir

0.17

 torino

0.15

tees

0.15

lant

0.15

ÙĦØ§ÙĦ

0.15

 newSize

0.14

erus

0.14

emiz

0.14

tl

0.14

Activations Density 0.014%

terms related to resource allocation

No Comments

No Known Activations