INDEX

Explanations

references to reading and processing written content

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

Ø§Ø¶

-0.15

pers

-0.14

izador

-0.14

ì§Ī

-0.14

èĪĮ

-0.13

clearfix

-0.13

Ø¶ÙĦ

-0.13

legg

-0.13

emy

-0.13

ia

-0.13

POSITIVE LOGITS

 reading

0.22

Reading

0.21

 Reading

0.19

éĺħè¯»

0.18

ãĥ¬ãĥĵ

0.17

 review

0.17

 reviewing

0.17

reading

0.17

-review

0.16

åľ¨çº¿éĺħè¯»

0.16

Activations Density 0.093%

references to reading and processing written content

No Comments

No Known Activations