INDEX

Explanations

instances of specific verbs related to observation and actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

pls

-0.15

achen

-0.15

PELL

-0.15

AFE

-0.15

itta

-0.14

itler

-0.14

itte

-0.14

uali

-0.14

anson

-0.14

arring

-0.14

POSITIVE LOGITS

 worden

0.37

 werden

0.28

 wird

0.23

 becoming

0.23

 wurde

0.23

Bec

0.21

 wordt

0.21

 become

0.20

 becomes

0.20

bec

0.19

Activations Density 0.012%

instances of specific verbs related to observation and actions

No Comments

No Known Activations