INDEX

Explanations

terms associated with educational credentials and programs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

was

-0.17

was

-0.16

 Ø¯Ø§Ø´Øª

-0.16

ãģĹãģ¾ãģĹãģŁ

-0.16

 hizo

-0.15

.isDefined

-0.15

 vardÄ±

-0.15

Was

-0.15

canf

-0.15

Ã¼nd

-0.15

POSITIVE LOGITS

 becomes

0.26

 blir

0.21

 begins

0.20

 become

0.20

 receives

0.19

 wordt

0.19

 decides

0.19

are

0.19

 gets

0.18

sp

0.17

Activations Density 0.426%

terms associated with educational credentials and programs

No Comments

No Known Activations