INDEX

Explanations

various references to games, educational assessments, and medical contexts

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

 stuff

-0.18

zione

-0.16

erk

-0.14

outing

-0.14

eger

-0.14

 Gaut

-0.14

legg

-0.13

 Stout

-0.13

jom

-0.13

 entries

-0.13

POSITIVE LOGITS

or

0.20

 nÃło

0.17

æĪĸ

0.17

 Ð¸Ð»Ð¸

0.16

(s

0.16

æĪĸèĢħ

0.16

stype

0.16

 ÛĮØ§

0.16

sWith

0.16

 Zaman

0.16

Activations Density 0.435%

various references to games, educational assessments, and medical contexts

No Comments

No Known Activations