INDEX

Explanations

expressions related to programming errors and exceptions

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

 mult

-0.16

tre

-0.15

dorf

-0.13

olie

-0.13

 multi

-0.13

bom

-0.13

 Mult

-0.13

 Gover

-0.13

incerely

-0.13

 Coord

-0.13

POSITIVE LOGITS

 Alone

0.16

 saja

0.15

ERV

0.15

-alone

0.15

 alone

0.15

ISTER

0.14

erta

0.14

 Simply

0.13

ATUS

0.13

 Solo

0.13

Activations Density 0.041%

expressions related to programming errors and exceptions

No Comments

No Known Activations