INDEX

Explanations

references to sources or resources that provide information or assistance

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

emed

-0.16

 Holl

-0.15

 flat

-0.15

engu

-0.15

aver

-0.15

ola

-0.15

oll

-0.14

trie

-0.14

avin

-0.14

ozo

-0.14

POSITIVE LOGITS

istrovstvÃŃ

0.16

ubern

0.16

LETE

0.15

.Listener

0.15

Ø´ÙĨ

0.15

/cop

0.15

oggler

0.14

innamon

0.14

atform

0.14

LOT

0.14

Activations Density 0.704%

references to sources or resources that provide information or assistance

No Comments

No Known Activations