INDEX

Explanations

references to individuals and their spoken opinions or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

icult

-0.16

 tell

-0.15

ulus

-0.15

zon

-0.15

amar

-0.15

elor

-0.14

 ÑģÐ¿ÑĢÐ¾Ñģ

-0.14

kit

-0.14

 telling

-0.14

POSITIVE LOGITS

.scalablytyped

0.23

continued

0.20

 continued

0.18

 explained

0.17

OTES

0.16

âĸį

0.16

 pointed

0.16

 Continued

0.16

 explan

0.16

Ð¾Ð·Ð²

0.15

Activations Density 0.047%

references to individuals and their spoken opinions or statements

No Comments

No Known Activations