INDEX

Explanations

repeated references to "each" and "every," indicating a focus on inclusivity or universal application

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

uese

-0.17

 çĶ°

-0.15

ile

-0.15

iew

-0.14

rim

-0.14

urger

-0.14

ÐºÐ¾Ð½

-0.14

erm

-0.14

argin

-0.13

Hat

-0.13

POSITIVE LOGITS

aris

0.17

Ð¼ÐµÑĩ

0.15

oeff

0.15

opak

0.14

.scalablytyped

0.14

Ø¶ÛĮ

0.14

each

0.14

ë§Īëĭ¤

0.14

itesse

0.14

Ð¸ÑĤÑĥ

0.13

Activations Density 0.036%

repeated references to "each" and "every," indicating a focus on inclusivity or universal application

No Comments

No Known Activations