INDEX

Explanations

references to additional individuals or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

shaw

-0.17

bable

-0.15

mong

-0.14

/token

-0.14

forth

-0.14

rico

-0.14

 thorough

-0.13

ned

-0.13

nable

-0.13

fort

-0.13

POSITIVE LOGITS

-than

0.23

 niÅ¼

0.21

world

0.19

 than

0.19

ials

0.19

wis

0.16

 kinds

0.15

ëĿ¼ëıĦ

0.15

ork

0.15

/new

0.15

Activations Density 0.109%

references to additional individuals or entities

No Comments

No Known Activations