INDEX

Explanations

references to notable characteristics and achievements of authors or creators

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 hous

-0.16

æ¼Ķ

-0.15

itech

-0.15

 æ¼Ķ

-0.14

mp

-0.14

 dancers

-0.14

bras

-0.13

ogg

-0.13

Dictionary

-0.13

POSITIVE LOGITS

 comics

0.43

 comic

0.39

 Comics

0.39

 graphic

0.35

comic

0.32

 Graphic

0.31

 strips

0.31

 strip

0.30

 Sequential

0.30

 manga

0.29

Activations Density 0.103%

references to notable characteristics and achievements of authors or creators

No Comments

No Known Activations