INDEX

Explanations

phrases related to analysis and expert consensus

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

imli

-0.17

st

-0.15

kn

-0.15

 attendance

-0.15

 Attendance

-0.14

kili

-0.14

 knot

-0.13

Sex

-0.13

 _____

-0.13

attendance

-0.13

POSITIVE LOGITS

iasi

0.16

fos

0.15

rrha

0.15

 Ð·Ð°Ð²Ð¸

0.14

ialis

0.14

swick

0.13

ayo

0.13

 porr

0.13

quina

0.13

ivent

0.13

Activations Density 0.206%

phrases related to analysis and expert consensus

No Comments

No Known Activations