INDEX

Explanations

phrases indicating safety and stability

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

fnlp/Llama-Scope-R1-Distill/800M-Slimpajama-0-OpenR1-Math-220k/L15R

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Hzfinfdu/SlimPajama-3B

Features

32,768

Data Type

float32

Hook Name

blocks.15.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

 unus

-0.14

 Pall

-0.14

 Details

-0.13

 backed

-0.13

 depending

-0.13

quist

-0.13

 Instant

-0.13

 Depending

-0.13

 Advanced

-0.13

 dependent

-0.13

POSITIVE LOGITS

æŃ£å¸¸

0.27

 normal

0.24

normal

0.20

.normal

0.19

ï¼Įæ²¡æľī

0.19

Normal

0.19

æĻ®éĢļ

0.18

ä¸įåŃĺåľ¨

0.17

.NoError

0.17

æ²¡æľī

0.16

Activations Density 0.102%

phrases indicating safety and stability

No Comments

No Known Activations