INDEX

Explanations

words related to medical or health conditions, particularly focusing on specific suffixes and terminologies

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

New Auto-Interp

Configuration

fnlp/Llama3_1-8B-Base-LXM-8x/Llama3_1-8B-Base-L22M-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

fffffff

-0.16

oute

-0.16

arov

-0.15

 competit

-0.15

urette

-0.15

Ces

-0.15

-0.14

ouri

-0.14

AKE

-0.14

ä¸ģ

-0.14

POSITIVE LOGITS

QUI

0.15

ulta

0.15

ventus

0.14

intendent

0.14

iba

0.14

DOB

0.14

amura

0.13

oto

0.13

uhn

0.13

à¤¿à¤¥

0.13

Activations Density 0.064%

words related to medical or health conditions, particularly focusing on specific suffixes and terminologies

No Comments

No Known Activations