INDEX

Explanations

words or phrases that indicate a ranking or rating

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ãģķãģĦ

-0.16

Ø¬Ø§

-0.15

BAB

-0.14

Ã¤ch

-0.14

hem

-0.14

scal

-0.14

Oak

-0.14

bab

-0.13

upy

-0.13

Ney

-0.13

POSITIVE LOGITS

 Persona

0.33

Persona

0.26

 persona

0.26

 golden

0.25

 Golden

0.25

 Person

0.24

persona

0.24

Golden

0.22

golden

0.22

 Velvet

0.21

Activations Density 0.000%

words or phrases that indicate a ranking or rating

No Comments

No Known Activations