INDEX

Explanations

phrases related to finding solutions or paths forward

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

enu

-0.15

unn

-0.15

urga

-0.14

nek

-0.14

Ø§ÙģØ¸

-0.14

ohn

-0.13

ley

-0.13

wa

-0.13

 Oswald

-0.13

BS

-0.13

POSITIVE LOGITS

 accent

0.15

getChild

0.14

Î³Îµ

0.14

ÑĢÐ¾ÑĤ

0.14

oric

0.14

owns

0.14

Laf

0.13

addock

0.13

finger

0.13

odia

0.13

Activations Density 0.021%

phrases related to finding solutions or paths forward

No Comments

No Known Activations