INDEX

Explanations

references to education, teaching, and training

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

 TÃ¼r

-0.16

lug

-0.15

udas

-0.15

 Templ

-0.15

Gew

-0.15

.crypto

-0.15

áº¡o

-0.14

bid

-0.14

icos

-0.14

į¨

-0.14

POSITIVE LOGITS

 intens

0.15

acam

0.15

kowski

0.14

 lessons

0.14

ForObject

0.14

PAD

0.14

 fold

0.14

 Fancy

0.14

cke

0.13

Activations Density 0.659%

references to education, teaching, and training

No Comments

No Known Activations