INDEX

Explanations

instances of various forms of the word "is."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

eron

-0.16

å´İ

-0.15

DBC

-0.15

¼

-0.15

uron

-0.14

ikers

-0.14

Å¸

-0.13

æ³ģ

-0.13

uncture

-0.13

ë

-0.13

POSITIVE LOGITS

Jar

0.14

VERR

0.14

Wag

0.14

chief

0.14

antro

0.14

ief

0.14

OTES

0.13

ãĤ·ãĤ¢

0.13

ores

0.13

Tun

0.13

Activations Density 0.002%

instances of various forms of the word "is."

No Comments

No Known Activations