INDEX

Explanations

phrases indicating comprehensive evaluations or assessments across various categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

alic

-0.16

uti

-0.15

kea

-0.15

 JAXBElement

-0.14

isch

-0.14

Ã¼mÃ¼ÅŁ

-0.14

osite

-0.14

ukt

-0.14

ihan

-0.14

 effect

-0.13

POSITIVE LOGITS

 spectrum

0.36

 entire

0.25

pectrum

0.24

 spectra

0.23

 Spectrum

0.23

 across

0.22

 Across

0.22

_spectrum

0.21

Across

0.21

æķ´ä¸ª

0.21

Activations Density 0.056%

phrases indicating comprehensive evaluations or assessments across various categories

No Comments

No Known Activations