INDEX

Explanations

phrases related to media releases and exclusive content

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

ICES

-0.17

 kinh

-0.15

.GetService

-0.15

Restart

-0.14

Î¿ÏĤ

-0.14

 Freund

-0.14

otron

-0.14

InBackground

-0.14

WithPath

-0.13

POSITIVE LOGITS

 exclusive

0.27

 exclusively

0.20

 exclus

0.19

exclusive

0.19

-exclusive

0.18

Exclusive

0.18

 never

0.17

 Exclusive

0.17

ervo

0.17

kening

0.16

Activations Density 0.096%

phrases related to media releases and exclusive content

No Comments

No Known Activations