INDEX

Explanations

references to experience and expertise in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

loth

-0.16

Ã²ng

-0.15

 ØŃØ§ÙģØ¸

-0.15

.dep

-0.15

à¤Ĥà¤ķ

-0.15

oca

-0.14

ikal

-0.14

 Spatial

-0.14

avic

-0.14

robe

-0.14

POSITIVE LOGITS

 gained

0.23

 dealing

0.17

zman

0.17

rtl

0.16

 acquired

0.16

 handling

0.15

in

0.15

/background

0.15

/k

0.15

0.14

Activations Density 0.033%

references to experience and expertise in various contexts

No Comments

No Known Activations