INDEX

Explanations

credit to contributors and editors in a work

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

umu

-0.17

han

-0.16

icari

-0.16

 Yates

-0.16

hani

-0.15

utory

-0.14

criptors

-0.14

ett

-0.14

otts

-0.14

uib

-0.14

POSITIVE LOGITS

_prefs

0.15

 ì§ĢëıĦ

0.15

ereal

0.15

ÐµÐ³ÑĢÐ°

0.15

metatable

0.14

íĭ°

0.14

ÑĢÑĥÑĩ

0.14

inde

0.14

0.13

 Pole

0.13

Activations Density 0.020%

credit to contributors and editors in a work

No Comments

No Known Activations