INDEX

Explanations

technical references to programming elements or structures

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

*>*

-0.17

anna

-0.15

vana

-0.15

Ã´i

-0.14

chia

-0.14

clid

-0.14

apro

-0.13

 firma

-0.13

ãģ¬

-0.13

Ð½ÑĸÐ¼

-0.13

POSITIVE LOGITS

Ã©ra

0.16

ãĥ¼ãĤ¹

0.15

 spur

0.15

oÄį

0.14

æŃ¢

0.14

islav

0.14

 ####

0.14

WithData

0.13

è¤

0.13

łéĻ¤

0.13

Activations Density 0.077%

technical references to programming elements or structures

No Comments

No Known Activations