INDEX

Explanations

phrases that indicate levels, particularly those around measurement or quality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

iaux

-0.16

rox

-0.16

zu

-0.16

erm

-0.15

ä»¥æĿ¥

-0.15

zw

-0.15

Ð³Ð¾Ð²

-0.15

ume

-0.15

ioc

-0.15

 pÃ¤

-0.14

POSITIVE LOGITS

ground

0.29

-average

0.27

 average

0.26

average

0.25

 ground

0.23

sea

0.22

 Average

0.21

 repro

0.20

 freezing

0.20

Average

0.20

Activations Density 0.023%

phrases that indicate levels, particularly those around measurement or quality

No Comments

No Known Activations