INDEX

Explanations

phrases that indicate improvement or enhancement in experiences or products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

Äł

-0.15

ropol

-0.14

/Dk

-0.14

iversit

-0.14

ItemAt

-0.14

etik

-0.14

ë°°

-0.13

rico

-0.13

ÃŃÅ¡

-0.12

ØªØ±ÙĨØª

-0.12

POSITIVE LOGITS

 everything

0.30

everything

0.25

 Everything

0.24

Everything

0.24

 tudo

0.24

ä¸ĢåĪĩ

0.22

 alles

0.19

 finally

0.17

 tutto

0.16

 migli

0.16

Activations Density 0.030%

phrases that indicate improvement or enhancement in experiences or products

No Comments

No Known Activations