INDEX

Explanations

references to options in product design

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

PTY

-0.17

 Rank

-0.16

ometr

-0.16

 rank

-0.16

 Hurt

-0.16

orum

-0.15

lash

-0.15

oten

-0.15

uja

-0.15

Rank

-0.15

POSITIVE LOGITS

 faucet

0.30

 fauc

0.30

 Fauc

0.29

 sink

0.26

Sink

0.24

 lever

0.23

 Sink

0.23

 sinks

0.23

 brass

0.23

 shower

0.22

Activations Density 0.013%

references to options in product design

No Comments

No Known Activations