INDEX

Explanations

phrases indicating built-in features or functionalities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

.Transactional

-0.16

/to

-0.15

igmat

-0.15

paque

-0.14

ARGIN

-0.14

Ø§Ø¦

-0.14

fce

-0.14

/people

-0.13

.isArray

-0.13

ichtig

-0.13

POSITIVE LOGITS

/embed

0.19

 sáºµn

0.17

omb

0.16

 Built

0.16

Built

0.15

uchar

0.15

alue

0.15

 built

0.15

built

0.15

-built

0.15

Activations Density 0.026%

phrases indicating built-in features or functionalities

No Comments

No Known Activations