INDEX

Explanations

references to hierarchical structures or classifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

®

-0.17

Î¼Î¿

-0.15

obble

-0.15

CORE

-0.14

tooltip

-0.14

 prÃ¡v

-0.14

boarding

-0.13

escription

-0.13

CONST

-0.13

uels

-0.13

POSITIVE LOGITS

archical

0.39

arch

0.33

archs

0.28

loom

0.22

ARCH

0.22

Ð°ÑĢÑħ

0.21

archy

0.20

onym

0.20

arcy

0.20

glyph

0.19

Activations Density 0.010%

references to hierarchical structures or classifications

No Comments

No Known Activations