INDEX

Explanations

information related to professional accomplishments and affiliations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

abin

-0.19

yt

-0.17

aket

-0.17

gend

-0.16

commons

-0.15

itu

-0.15

aci

-0.15

akan

-0.15

fit

-0.14

 sint

-0.14

POSITIVE LOGITS

 Society

0.25

 Toast

0.24

Toast

0.23

 society

0.22

 American

0.22

Eta

0.22

Tau

0.22

 Lambda

0.21

Phi

0.21

 Mens

0.21

Activations Density 0.112%

information related to professional accomplishments and affiliations

No Comments

No Known Activations