INDEX

Explanations

terms related to delays and postponements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.15.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

ibaba

-0.17

rot

-0.14

en

-0.14

DataService

-0.14

 trapped

-0.14

_cv

-0.14

acyj

-0.13

icolor

-0.13

alu

-0.13

POSITIVE LOGITS

 postponed

0.29

 postpon

0.28

 postpone

0.24

defer

0.21

 delay

0.19

 delaying

0.19

 defer

0.19

CANCEL

0.19

 scheduled

0.19

 Delay

0.19

Activations Density 0.114%

terms related to delays and postponements

No Comments

No Known Activations