INDEX

Explanations

instructions or statements indicating actions to be taken

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

ifice

-0.19

lation

-0.17

cre

-0.16

rias

-0.16

cribe

-0.16

pu

-0.15

sp

-0.15

wel

-0.15

shan

-0.15

uments

-0.15

POSITIVE LOGITS

 whom

0.24

xic

0.23

onces

0.21

ilet

0.20

emailer

0.19

getter

0.19

infinity

0.19

 Hell

0.19

Ãłn

0.19

owo

0.18

Activations Density 0.048%

instructions or statements indicating actions to be taken

No Comments

No Known Activations