INDEX

Explanations

phrases indicating benefits or advantages related to various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

orado

-0.17

onth

-0.15

onation

-0.14

itin

-0.14

xF

-0.14

getic

-0.13

vr

-0.13

prak

-0.13

 nation

-0.13

ÑĬ

-0.13

POSITIVE LOGITS

 having

0.21

 Having

0.17

having

0.17

Having

0.17

eref

0.15

uries

0.15

herent

0.14

taj

0.14

ime

0.14

ere

0.14

Activations Density 0.076%

phrases indicating benefits or advantages related to various topics

No Comments

No Known Activations