INDEX

Explanations

terms related to engineering and engineering disciplines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

rahim

-0.17

ication

-0.17

GGLE

-0.17

istence

-0.17

deniz

-0.16

anders

-0.16

 WINAPI

-0.15

_codec

-0.15

ÐľÐŀ

-0.15

ppy

-0.14

POSITIVE LOGITS

ered

0.32

ering

0.25

å¸«

0.20

 feats

0.20

å¸Ī

0.19

eer

0.19

ERING

0.19

/arch

0.19

/design

0.18

/sc

0.18

Activations Density 0.014%

terms related to engineering and engineering disciplines

No Comments

No Known Activations