INDEX

Explanations

terms related to software functionality and user interface features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

akis

-0.16

 VÅ©

-0.15

VML

-0.15

eness

-0.15

ãĤįãģĨ

-0.14

analytics

-0.14

_HC

-0.14

.Encoding

-0.14

-assets

-0.13

Injected

-0.13

POSITIVE LOGITS

 Wizards

0.23

 stored

0.20

 operator

0.20

 menu

0.19

 graphic

0.18

 Windows

0.18

 wizards

0.18

 windows

0.18

 saved

0.17

 Stored

0.17

Activations Density 0.046%

terms related to software functionality and user interface features

No Comments

No Known Activations