INDEX

Explanations

superlative adjectives and phrases that indicate the highest level of quality or status

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

Ú¯ÛĮØ±ÛĮ

-0.16

.appspot

-0.15

_Exception

-0.14

ÑĭÐ¿

-0.14

nip

-0.14

ihan

-0.14

inen

-0.14

íĺľ

-0.13

ÏĥÏĦÎ¿

-0.13

vÄĽdom

-0.13

POSITIVE LOGITS

McM

0.15

 McKay

0.14

ions

0.14

 Behind

0.14

ucu

0.14

 lamb

0.14

ÙħØ§ÙĦ

0.14

most

0.14

 spit

0.14

ãĥ¼ãĥĹ

0.14

Activations Density 0.094%

superlative adjectives and phrases that indicate the highest level of quality or status

No Comments

No Known Activations