INDEX

Explanations

phrases related to examining or assessing something closely

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 Appearance

-0.36

 appearance

-0.35

Appearance

-0.33

appearance

-0.32

 appearances

-0.32

 appearing

-0.27

appear

-0.27

 appears

-0.26

 Appears

-0.26

 appear

-0.25

POSITIVE LOGITS

 look

0.62

look

0.50

 Look

0.49

Look

0.45

 LOOK

0.43

_look

0.42

.look

0.42

 looked

0.41

 looks

0.41

 looking

0.41

Activations Density 0.068%

phrases related to examining or assessing something closely

No Comments

No Known Activations