INDEX

Explanations

references to enjoyment and positive human experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

sto

-0.15

piar

-0.15

interopRequire

-0.15

udiante

-0.15

 mess

-0.14

Ļ

-0.14

Ĭ

-0.14

oga

-0.14

Ä±klÄ±

-0.13

POSITIVE LOGITS

@a

0.15

ãĥ©ãĤ¯

0.15

Ð¸ÑĤÐ¸Ð²

0.15

hed

0.14

ensing

0.14

.localtime

0.14

ursal

0.13

åĲ

0.13

emer

0.13

 tiener

0.13

Activations Density 0.003%

references to enjoyment and positive human experiences

No Comments

No Known Activations