INDEX

Explanations

references to matching and block structures in code or mathematical contexts

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

queda

-0.16

usk

-0.14

CP

-0.14

usp

-0.13

voir

-0.13

late

-0.13

 eens

-0.13

 cons

-0.13

ero

-0.13

eparator

-0.13

POSITIVE LOGITS

dbe

0.15

ĶåĽŀ

0.15

 Grove

0.14

èģ¯

0.14

nell

0.14

amer

0.14

èģĶ

0.13

urope

0.13

ãĥĬãĥ¼

0.13

Cin

0.13

Activations Density 0.102%

references to matching and block structures in code or mathematical contexts

No Comments

No Known Activations