INDEX

Explanations

code block indicators and formatting in markup languages

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ennie

-0.17

vertiser

-0.15

vert

-0.15

 Lace

-0.15

fts

-0.14

 Castillo

-0.14

ron

-0.14

 nylon

-0.13

essler

-0.13

 Lester

-0.13

POSITIVE LOGITS

`↵

0.19

uhn

0.19

ownik

0.17

uze

0.16

omics

0.16

 å°ĳ

0.16

ForRow

0.15

atre

0.15

0.14

eland

0.14

Activations Density 0.003%

code block indicators and formatting in markup languages

No Comments

No Known Activations