INDEX

Explanations

specific terms and symbols often related to digital content and communication

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 DBHelper

-0.13

 shortened

-0.13

 unity

-0.13

à¹Ħà¸«à¸Ļ

-0.13

unity

-0.13

-caret

-0.13

Reporting

-0.13

wb

-0.13

.YesNo

-0.13

 Sever

-0.13

POSITIVE LOGITS

lag

0.17

 concurrent

0.15

 tard

0.15

 Brilliant

0.15

 complement

0.15

preced

0.15

 corresponding

0.15

 stagn

0.14

 compliment

0.14

compet

0.14

Activations Density 0.032%

specific terms and symbols often related to digital content and communication

No Comments

No Known Activations