INDEX

Explanations

phrases that indicate methodologies and findings in research

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 Haut

-0.17

dbg

-0.16

ideo

-0.15

ests

-0.15

nil

-0.15

bak

-0.15

ddy

-0.14

rowser

-0.14

leet

-0.14

amas

-0.14

POSITIVE LOGITS

 tract

0.23

straight

0.20

 comput

0.17

æĮº

0.17

tract

0.17

 concept

0.17

 Ð¿Ð»Ð¾Ð´

0.16

 attractive

0.16

 attracting

0.16

Straight

0.16

Activations Density 0.092%

phrases that indicate methodologies and findings in research

No Comments

No Known Activations