INDEX

Explanations

connections between scientific concepts and popular understanding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

olah

-0.15

ailand

-0.15

estre

-0.14

otti

-0.14

itag

-0.14

aly

-0.14

oreal

-0.14

ÃĹ↵↵

-0.14

arus

-0.13

.kotlin

-0.13

POSITIVE LOGITS

but

0.30

but

0.26

 nhÆ°ng

0.25

ï¼Įä½Ĩ

0.24

 zwar

0.23

 Ð½Ð¾

0.22

 pero

0.21

ãģłãģĮ

0.21

 ÙĦÙĥÙĨ

0.20

 somehow

0.20

Activations Density 0.128%

connections between scientific concepts and popular understanding

No Comments

No Known Activations