INDEX

Explanations

occurrences of temporal phrases and conditions related to events or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

oppable

-0.15

ropri

-0.14

akedirs

-0.14

idunt

-0.14

æļ

-0.14

chwitz

-0.13

lien

-0.13

ä¸Īå¤«

-0.13

á¿Ĩ

-0.13

EMA

-0.13

POSITIVE LOGITS

 somebody

0.24

 someone

0.22

run

0.21

 user

0.19

 ever

0.18

ran

0.18

Someone

0.17

æŁĲ

0.17

-ever

0.17

 Somebody

0.17

Activations Density 0.186%

occurrences of temporal phrases and conditions related to events or actions

No Comments

No Known Activations