INDEX

Explanations

references to titles, wins, and achievements in competitive contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.15.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

oten

-0.15

TestFixture

-0.14

dition

-0.14

avl

-0.14

 Mick

-0.14

ersive

-0.13

erge

-0.13

 schedule

-0.13

 Grove

-0.13

icultural

-0.13

POSITIVE LOGITS

 hearts

0.20

 battles

0.20

 battle

0.18

Î±Î½Î¬

0.18

 Battles

0.17

nable

0.16

 Hearts

0.16

battle

0.16

 Battle

0.15

 Ð¿ÑĥÑĤÐµÐ¼

0.15

Activations Density 0.103%

references to titles, wins, and achievements in competitive contexts

No Comments

No Known Activations