INDEX

Explanations

references to quizzes and game formats

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

abus

-0.16

elli

-0.16

ement

-0.15

alink

-0.15

ifu

-0.15

ÐĶÐļ

-0.15

chaft

-0.15

ncy

-0.15

akis

-0.14

åĳ¨æľŁ

-0.14

POSITIVE LOGITS

à¥įà¤¤à¤ķ

0.17

 ResourceType

0.16

/photo

0.15

illac

0.15

 warehouse

0.14

AbsolutePath

0.14

ebra

0.14

lington

0.14

zier

0.14

ront

0.14

Activations Density 0.026%

references to quizzes and game formats

No Known Activations