INDEX

Explanations

references to global or world-class entities and achievements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

anne

-0.18

aken

-0.15

aller

-0.15

ampa

-0.15

antar

-0.15

 Internacional

-0.15

or

-0.15

aster

-0.14

chen

-0.14

 Gund

-0.14

POSITIVE LOGITS

-class

0.23

-ren

0.23

-wide

0.22

-leading

0.21

 wide

0.21

Wide

0.21

wide

0.19

 premiere

0.19

-record

0.19

 record

0.18

Activations Density 0.029%

references to global or world-class entities and achievements

No Comments

No Known Activations