INDEX

Explanations

references to medical recommendations and prescriptions, particularly in relation to the rules and behaviors of doctors and clinics

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

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

owie

-0.16

wor

-0.15

tile

-0.15

ÏĢÎ¿ÏĦÎµ

-0.14

iges

-0.14

 Guarantee

-0.13

 gÃ¶l

-0.13

iero

-0.13

CRY

-0.13

 Ã¼r

-0.13

POSITIVE LOGITS

 issuing

0.20

 branch

0.16

 charging

0.16

kee

0.15

 branches

0.15

 granting

0.14

 giving

0.14

cke

0.14

 branching

0.14

é¢

0.14

Activations Density 0.075%

references to medical recommendations and prescriptions, particularly in relation to the rules and behaviors of doctors and clinics

No Comments

No Known Activations