INDEX

Explanations

place and event names, especially in entertainment and politics

oai_token-act-pair · gemini-2.0-flash

New Auto-Interp

Configuration

fnlp/Llama-Scope-R1-Distill/400M-Slimpajama-400M-OpenR1-Math-220k/L11R

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Hzfinfdu/SlimPajama-3B and open-r1/OpenR1-Math-220k

Features

32,768

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vs

-0.08

 Â«

-0.07

 (Â«

-0.07

 special

-0.07

 micro

-0.07

Ag

-0.07

 home

-0.07

out

-0.06

 ourselves

-0.06

POSITIVE LOGITS

inÃ§

0.07

OffsetTable

0.07

 má»±c

0.06

ãĥ£

0.06

à¥Ĥà¤¡

0.06

:NO

0.06

Ø§ÛĮØ²

0.06

Ð½Ð¾ÑģÑı

0.06

çİ°åľº

0.06

ernote

0.06

Activations Density 0.209%

place and event names, especially in entertainment and politics

No Comments

No Known Activations