INDEX

Explanations

phrases indicating purpose or intention

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

ÑĢÐµÐº

-0.16

inate

-0.15

-0.14

ser

-0.14

kk

-0.14

UPDATED

-0.14

uh

-0.13

bid

-0.13

oi

-0.13

POSITIVE LOGITS

 toward

0.35

 towards

0.32

Towards

0.24

 Towards

0.23

 hacia

0.21

owards

0.21

åĲĳ

0.20

Tow

0.20

 verso

0.18

adero

0.17

Activations Density 0.037%

phrases indicating purpose or intention

No Comments

No Known Activations