INDEX

Explanations

numerical values or figures in text

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

izar

-0.16

Ð¾ÑĢÐ°Ð·

-0.15

 July

-0.15

enza

-0.15

 instrument

-0.14

 October

-0.14

 Julius

-0.14

etto

-0.14

ropped

-0.14

ioni

-0.14

POSITIVE LOGITS

 ValueEventListener

0.15

 chÃ¢n

0.15

Ã¡li

0.15

WSC

0.15

endar

0.15

orado

0.15

ENCH

0.14

íĢ

0.14

ÑĢÐ¾Ð³

0.14

Picture

0.14

Activations Density 0.031%

numerical values or figures in text

No Comments

No Known Activations