INDEX

Explanations

concepts related to physical training and instructional scenarios

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 consumer

-0.18

pell

-0.15

nbr

-0.15

 epis

-0.14

 blunt

-0.14

 shrink

-0.14

 community

-0.14

 document

-0.14

 public

-0.14

POSITIVE LOGITS

 Technique

0.24

 technique

0.22

 execute

0.17

execution

0.17

 executes

0.17

 execution

0.17

 executing

0.17

Execution

0.16

_execution

0.16

acements

0.16

Activations Density 0.348%

concepts related to physical training and instructional scenarios

No Comments

No Known Activations