INDEX

Explanations

instances and examples in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 indeed

-0.15

ãģ¾ãģŁ

-0.14

ija

-0.14

para

-0.13

azar

-0.13

æ¡ĥ

-0.13

_exceptions

-0.13

 Ð·Ð¾ÐºÑĢÐµÐ¼Ð°

-0.13

Î¼ÏīÏĤ

-0.13

ico

-0.13

POSITIVE LOGITS

 sake

0.28

 purposes

0.23

0.20

orz

0.16

:↵

0.16

pillar

0.16

èĢĮ

0.15

ãģĪãģ°

0.15

many

0.15

 kdyÅ¾

0.15

Activations Density 0.031%

instances and examples in the text

No Comments

No Known Activations