INDEX

Explanations

technical terms and specifications related to automotive design and engineering

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 Gazette

-0.16

old

-0.15

å¤ªéĥİ

-0.15

agher

-0.14

 often

-0.14

 Sund

-0.14

 Gamb

-0.14

çĭĲ

-0.14

_pb

-0.14

CHA

-0.14

POSITIVE LOGITS

ounge

0.16

Ä±rak

0.15

serial

0.15

 emoc

0.15

 Serial

0.15

orde

0.15

 Concept

0.15

 Downs

0.15

 Anth

0.15

umba

0.14

Activations Density 0.013%

technical terms and specifications related to automotive design and engineering

No Comments

No Known Activations