INDEX

Explanations

inquiries and opinions from the audience

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

weetalert

-0.15

amburger

-0.15

ä¹ĥ

-0.15

ightly

-0.14

weit

-0.14

åħ·

-0.14

ingo

-0.14

à¥įà¤¦à¤°

-0.14

PushMatrix

-0.13

Î¹

-0.13

POSITIVE LOGITS

 think

0.52

 thinks

0.48

 Think

0.47

Think

0.45

 thoughts

0.43

think

0.41

 thinking

0.40

 THINK

0.40

 thought

0.39

 Thoughts

0.38

Activations Density 0.051%

inquiries and opinions from the audience

No Comments

No Known Activations