INDEX

Explanations

significant changes or events, particularly in a context of evaluation or assessment

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

stre

-0.15

fore

-0.15

 Mond

-0.15

 Perr

-0.15

runs

-0.14

 fore

-0.14

 extent

-0.14

é½Ĳ

-0.14

omm

-0.14

ailles

-0.13

POSITIVE LOGITS

 resulted

0.32

 produces

0.30

 produce

0.30

å¯¼èĩ´

0.28

 producing

0.28

 Produ

0.26

produce

0.25

 Produce

0.23

produ

0.23

 generates

0.23

Activations Density 0.028%

significant changes or events, particularly in a context of evaluation or assessment

No Comments

No Known Activations