INDEX

Explanations

phrases indicating condition or quality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

Zug

-0.16

rike

-0.16

ucha

-0.14

ãĤ¤ãĥĪ

-0.14

 Nack

-0.14

ubern

-0.14

Digit

-0.14

icit

-0.14

DFS

-0.14

AGO

-0.13

POSITIVE LOGITS

elik

0.15

ureau

0.14

à¸ªà¸°

0.14

alue

0.13

aty

0.13

Story

0.13

 Mand

0.13

ivol

0.13

ConnectionFactory

0.13

Activations Density 0.032%

phrases indicating condition or quality

No Comments

No Known Activations