INDEX

Explanations

phrases expressing quantity or degree of emphasis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

.EVT

-0.15

ÙĪØ¨Ø©

-0.14

strup

-0.14

iyah

-0.14

PF

-0.14

alia

-0.14

ARIO

-0.14

CCCCCC

-0.14

ÑħÐ¾Ð´Ð¸Ð¼

-0.14

.opend

-0.14

POSITIVE LOGITS

Ð»Ð¸Ð½Ð¸

0.15

elson

0.14

 tato

0.14

Ð»ÐºÐ¸

0.14

yards

0.14

veau

0.14

owell

0.13

"urls

0.13

ordes

0.13

Balancer

0.13

Activations Density 0.014%

phrases expressing quantity or degree of emphasis

No Comments

No Known Activations