INDEX

Explanations

phrases related to self-examination and contemplation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

deen

-0.17

ÑĢÐµÐ±

-0.16

ãģ°

-0.16

erken

-0.15

elan

-0.15

uur

-0.15

frau

-0.14

ereotype

-0.14

Ã¢y

-0.14

"url

-0.14

POSITIVE LOGITS

0.17

0.16

arem

0.16

 reflect

0.15

 reflection

0.15

reflect

0.15

owl

0.15

 thought

0.14

iveness

0.14

ľ

0.14

Activations Density 0.027%

phrases related to self-examination and contemplation

No Comments

No Known Activations