INDEX

Explanations

numerical values, particularly those representing dates and fiscal years

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

 seventh

-0.16

elay

-0.15

 seven

-0.15

 Jana

-0.15

cers

-0.15

.jupiter

-0.15

olie

-0.14

ä¸ĥ

-0.14

plet

-0.14

¸ı

-0.14

POSITIVE LOGITS

0.27

four

0.24

-four

0.24

 FOUR

0.21

 Fourth

0.21

 fourth

0.20

 Four

0.20

uzzi

0.20

 four

0.20

 dÃ¶rt

0.20

Activations Density 0.159%

numerical values, particularly those representing dates and fiscal years

No Comments

No Known Activations