INDEX

Explanations

words related to collaboration and partnerships in various projects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

Ð¾Ð±Ð¸

-0.16

afort

-0.15

ifest

-0.14

isay

-0.14

ocha

-0.14

åİļ

-0.14

edics

-0.13

iesta

-0.13

edi

-0.13

UNS

-0.13

POSITIVE LOGITS

 alike

0.28

 themselves

0.23

 together

0.23

 respectively

0.23

åħ±åĲĮ

0.20

 Together

0.20

 respective

0.18

gether

0.18

joint

0.17

Together

0.17

Activations Density 0.190%

words related to collaboration and partnerships in various projects

No Comments

No Known Activations