INDEX

Explanations

concepts related to strategies, systems, and structures in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

 Lowest

-0.15

ported

-0.14

 relaxed

-0.14

auer

-0.13

 Loaded

-0.13

ert

-0.13

 nÃ³i

-0.13

 budding

-0.13

ugh

-0.13

 ÙģØ±Ø§ÙĩÙħ

-0.13

POSITIVE LOGITS

 whose

0.31

 that

0.30

that

0.28

whose

0.26

 which

0.23

_THAT

0.21

	that

0.20

_that

0.20

 Å¼e

0.19

 daÃŁ

0.17

Activations Density 1.029%

concepts related to strategies, systems, and structures in various contexts

No Comments

No Known Activations