INDEX

Explanations

HTML or web navigation elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

iales

-0.16

Bez

-0.15

 progresses

-0.15

 mÃ¸

-0.15

 Animalia

-0.14

wald

-0.14

enie

-0.14

aÅ¼

-0.13

Ð¾Ð¼ÐµÐ½

-0.13

POSITIVE LOGITS

 next

0.15

ano

0.14

aspect

0.14

unte

0.14

next

0.14

BM

0.14

MI

0.14

èįĴ

0.14

MT

0.13

../

0.13

Activations Density 0.015%

HTML or web navigation elements

No Comments

No Known Activations