INDEX

Explanations

phrases that reference various professional associations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

apan

-0.15

Æ¡

-0.14

Ø³Ø§Ø¨

-0.14

isches

-0.14

wu

-0.14

(Font

-0.14

upal

-0.14

enti

-0.13

á»ķ

-0.13

Éµ

-0.13

POSITIVE LOGITS

.scalablytyped

0.16

 Chiefs

0.15

cord

0.14

hol

0.14

 cord

0.14

ROP

0.14

eed

0.14

 Computing

0.14

 remar

0.13

ofday

0.13

Activations Density 0.013%

phrases that reference various professional associations

No Comments

No Known Activations