INDEX

Explanations

numerical values or statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

puter

-0.21

CEL

-0.16

.cf

-0.16

conomy

-0.15

esper

-0.15

Ã´ng

-0.13

-turned

-0.13

ÑĢÐ°Ð²Ð¸

-0.13

[ip

-0.13

oulos

-0.13

POSITIVE LOGITS

uzey

0.20

-REAL

0.15

aka

0.14

 toto

0.14

Bid

0.14

 Grimm

0.14

 compensated

0.13

luk

0.13

LOAT

0.13

issen

0.13

Activations Density 0.002%

numerical values or statistics

No Comments

No Known Activations