INDEX

Explanations

references to sports competitions and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

 surre

-0.14

 Peyton

-0.14

 McGregor

-0.14

Æ°á»Ľi

-0.13

 Simpson

-0.13

Å¾enÃŃ

-0.13

-images

-0.13

ebi

-0.13

ultip

-0.13

Ð´Ñı

-0.13

POSITIVE LOGITS

ucher

0.16

 Insecta

0.15

.strict

0.15

mada

0.15

é¬

0.15

à¸±à¸ĩà¸Ħ

0.15

arshal

0.14

azon

0.14

éĻĲ

0.14

EdgeInsets

0.14

Activations Density 0.012%

references to sports competitions and titles

No Comments

No Known Activations