INDEX

Explanations

instructions related to technical procedures or troubleshooting steps

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

 surre

-0.14

 Quiz

-0.14

á»ĭ

-0.14

Î¼Î¬

-0.14

quiz

-0.14

eniable

-0.14

portun

-0.14

usan

-0.14

endoza

-0.13

quete

-0.13

POSITIVE LOGITS

 instructions

0.60

 instruction

0.50

instructions

0.48

 Instructions

0.46

 steps

0.46

Instructions

0.43

 step

0.42

 directions

0.42

instruction

0.38

_instructions

0.38

Activations Density 0.278%

instructions related to technical procedures or troubleshooting steps

No Comments

No Known Activations