INDEX

Explanations

words related to various topics or items discussed in a document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 Washington

-0.16

 Schumer

-0.15

Washington

-0.14

 Oprah

-0.14

 manufacturers

-0.14

 golf

-0.14

MLB

-0.14

 Oval

-0.13

 cref

-0.13

 Chen

-0.13

POSITIVE LOGITS

 Pirate

0.47

 Pirates

0.38

 pirate

0.38

 pirates

0.34

Pir

0.28

pir

0.26

 Party

0.26

pir

0.24

Party

0.23

 party

0.23

Activations Density 0.004%

words related to various topics or items discussed in a document

No Comments

No Known Activations