INDEX

Explanations

references to the concept of "flipping" or "flip-flops."

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

estro

-0.17

ughter

-0.16

icated

-0.15

iyat

-0.15

 fisse

-0.15

.fig

-0.14

hurst

-0.14

FIG

-0.14

naments

-0.14

à¤¶à¤¨

-0.14

POSITIVE LOGITS

 flop

0.33

per

0.30

 flips

0.29

 flip

0.29

flo

0.29

 Flip

0.28

kart

0.27

flip

0.27

Flo

0.26

Flip

0.26

Activations Density 0.007%

references to the concept of "flipping" or "flip-flops."

No Comments

No Known Activations