INDEX

Explanations

symbols and punctuation, specifically focusing on commas and periods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

zers

-0.16

fty

-0.15

odos

-0.15

Äĥn

-0.14

HEME

-0.14

illisecond

-0.14

à¤¾à¤ĩà¤²

-0.14

bis

-0.14

Ã¨o

-0.13

 DÄĽ

-0.13

POSITIVE LOGITS

íĬ¼

0.15

iegel

0.14

 Poly

0.14

ÅĻÃŃzenÃŃ

0.14

@brief

0.14

ertools

0.14

 Oral

0.13

 Expires

0.13

ector

0.13

uilder

0.13

Activations Density 0.342%

symbols and punctuation, specifically focusing on commas and periods

No Comments

No Known Activations