INDEX

Explanations

terms related to programming concepts or code structure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

Hav

-0.15

peg

-0.14

 Kong

-0.14

 fossil

-0.14

ajar

-0.14

.spatial

-0.14

ang

-0.14

PTH

-0.13

 editorial

-0.13

áº¡t

-0.13

POSITIVE LOGITS

 ÑĤÐ²Ð°

0.18

âĦĸâĦĸ

0.16

oeff

0.16

 addCriterion

0.16

 carte

0.16

uco

0.15

ÐºÑĥÑĢ

0.14

~-~-~-~-

0.14

rada

0.14

tura

0.14

Activations Density 0.076%

terms related to programming concepts or code structure

No Comments

No Known Activations