INDEX

Explanations

references to conferences and organized events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ÙĬØ±Ø§

-0.16

irie

-0.15

ccount

-0.13

Trad

-0.13

anova

-0.13

AMPLE

-0.13

olia

-0.12

aily

-0.12

íĨµ

-0.12

 Episodes

-0.12

POSITIVE LOGITS

sym

0.51

 conference

0.48

 forum

0.42

 Conference

0.38

 summit

0.38

conference

0.38

 Symposium

0.38

sym

0.37

 conferences

0.35

Sym

0.35

Activations Density 0.227%

references to conferences and organized events

No Comments

No Known Activations