INDEX

Explanations

phrases indicating relationships or connections

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

 jenter

-0.16

 Bail

-0.15

amus

-0.15

_IW

-0.15

 lekker

-0.14

íĴį

-0.14

amet

-0.14

amodel

-0.14

aurus

-0.14

à¸ļà¸ģ

-0.13

POSITIVE LOGITS

 Germ

0.17

ger

0.16

åħ±

0.16

 COMM

0.15

 Share

0.15

share

0.15

-share

0.15

ger

0.14

 share

0.14

rob

0.14

Activations Density 0.019%

phrases indicating relationships or connections

No Comments

No Known Activations