INDEX

Explanations

terms related to structural attributes and organization in documents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

Ð°ÑĢÑĩ

-0.17

tee

-0.17

Ð°ÑĢ

-0.16

 arcs

-0.16

 dist

-0.15

AR

-0.15

ARING

-0.15

Bad

-0.14

Tun

-0.14

aring

-0.14

POSITIVE LOGITS

oya

0.15

elps

0.15

ipes

0.15

rub

0.15

rub

0.15

oy

0.15

 rubbing

0.15

FRING

0.14

ong

0.14

opsis

0.14

Activations Density 0.021%

terms related to structural attributes and organization in documents

No Comments

No Known Activations