INDEX

Explanations

references to specific years and ages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

.wik

-0.16

elden

-0.15

ojis

-0.15

ÐµÐ½ÑĮ

-0.15

oop

-0.15

yor

-0.14

opper

-0.14

inka

-0.14

entin

-0.14

anuts

-0.14

POSITIVE LOGITS

ning

0.17

RelativeTo

0.15

Ã³Å¼

0.15

annies

0.15

0.14

ä»½

0.14

ed

0.14

ÙĨØ§ÙħÙĩ

0.14

uzu

0.14

ë§Īëĭ¤

0.13

Activations Density 0.021%

references to specific years and ages

No Comments

No Known Activations