INDEX

Explanations

information related to data collection and analysis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

Ø±ÙĪØª

-0.15

ISTA

-0.15

uggy

-0.14

 Shine

-0.14

 acceler

-0.14

 shine

-0.14

 Rare

-0.14

 homes

-0.14

 Slee

-0.14

 pert

-0.14

POSITIVE LOGITS

abase

0.22

oidal

0.17

arra

0.16

odate

0.15

orizontal

0.15

 information

0.15

-information

0.15

AF

0.15

.createStatement

0.15

strup

0.14

Activations Density 0.173%

information related to data collection and analysis

No Comments

No Known Activations