INDEX

Explanations

phrases that indicate personal reflections and opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

deaux

-0.15

letcher

-0.15

ÑĪÐ¸Ð½

-0.14

acked

-0.14

 McKay

-0.14

 mass

-0.14

Ð±Ð¾ÑĢÐ°

-0.13

zb

-0.13

POSITIVE LOGITS

 something

0.39

 ones

0.37

something

0.35

Something

0.33

 Something

0.31

omething

0.29

 areas

0.21

 Ones

0.21

ones

0.20

ä¸Ģç§į

0.20

Activations Density 0.220%

phrases that indicate personal reflections and opinions

No Comments

No Known Activations