INDEX

Explanations

elements related to educational and professional accomplishments

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

uvo

-0.18

dÃ¼k

-0.17

imer

-0.15

bane

-0.15

afx

-0.15

edith

-0.15

Kro

-0.14

orial

-0.14

otec

-0.14

ereg

-0.14

POSITIVE LOGITS

amm

0.14

riba

0.14

etz

0.13

ÙĬÙĬÙĨ

0.13

IEW

0.13

anche

0.13

itel

0.13

ph

0.13

nev

0.13

 prize

0.13

Activations Density 0.223%

elements related to educational and professional accomplishments

No Comments

No Known Activations