INDEX

Explanations

phrases indicating discovery or inquiry

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

åĥį

-0.15

uset

-0.15

upy

-0.15

Ø§ÙģØª

-0.14

ropol

-0.14

ÅĻ

-0.14

DOWN

-0.14

ruz

-0.13

aim

-0.13

Į¨

-0.13

POSITIVE LOGITS

out

0.89

-out

0.54

out

0.51

åĩº

0.46

.out

0.43

out

0.41

Out

0.41

_out

0.41

(out

0.40

 Ð¸Ð·

0.37

Activations Density 0.041%

phrases indicating discovery or inquiry

No Comments

No Known Activations