INDEX

Explanations

punctuation and symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 dames

-0.13

 sund

-0.13

ESL

-0.13

gon

-0.12

ë§¥

-0.12

Ø§Ø¨Ø·

-0.12

 denomin

-0.12

®

-0.12

åľŃ

-0.12

-Mart

-0.12

POSITIVE LOGITS

awi

0.15

cken

0.14

YK

0.13

ogg

0.13

ÃŃas

0.13

fty

0.13

ï»¿using

0.13

osit

0.12

owering

0.12

ioneer

0.12

Activations Density 0.065%

punctuation and symbols

No Comments

No Known Activations