INDEX

Explanations

various types of metadata and organization labels in a document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

ullan

-0.14

Tin

-0.14

lient

-0.14

kud

-0.13

erville

-0.13

Gaw

-0.13

 Narr

-0.13

_Tool

-0.13

iek

-0.13

uevo

-0.13

POSITIVE LOGITS

chia

0.15

ichni

0.15

 hete

0.15

orget

0.14

hend

0.14

Gram

0.14

OGLE

0.14

burger

0.13

ãĥ¼ãĥ

0.13

 ----------------------------------------------------------------------↵

0.13

Activations Density 0.068%

various types of metadata and organization labels in a document

No Comments

No Known Activations