INDEX

Explanations

references to inner and outer spaces or concepts, particularly related to community or urban environments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

iec

-0.19

andan

-0.18

sin

-0.18

sdale

-0.17

ron

-0.16

rus

-0.16

scar

-0.15

services

-0.15

entifier

-0.15

istring

-0.15

POSITIVE LOGITS

most

0.47

halb

0.34

-inner

0.28

MOST

0.27

-most

0.26

(inner

0.24

 workings

0.23

 most

0.22

-city

0.22

wear

0.22

Activations Density 0.016%

references to inner and outer spaces or concepts, particularly related to community or urban environments

No Comments

No Known Activations