INDEX

Explanations

references to historical events and their causes or consequences

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

 folks

-0.15

 Universe

-0.15

endency

-0.15

 Loud

-0.14

usses

-0.14

edik

-0.14

Cla

-0.14

Fol

-0.14

Mac

-0.14

ktor

-0.14

POSITIVE LOGITS

maj

0.16

acÃŃ

0.15

ObjectContext

0.15

anine

0.14

uki

0.14

ogany

0.14

odiac

0.14

_inp

0.13

åĩ

0.13

olas

0.13

Activations Density 1.233%

references to historical events and their causes or consequences

No Comments

No Known Activations