INDEX

Explanations

references to names or titles, particularly within a specific context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

tu

-0.16

okus

-0.15

ÏĥÎ¿Ïħ

-0.15

lut

-0.14

çº¯

-0.14

uchs

-0.14

ÑĩÐµÑģÐºÐ¸

-0.14

ardu

-0.14

nio

-0.14

 Slut

-0.14

POSITIVE LOGITS

aul

0.17

iaz

0.16

 requestOptions

0.15

 chung

0.15

æ²¢

0.15

edia

0.14

sthrough

0.13

.Mask

0.13

Mol

0.13

.NODE

0.13

Activations Density 0.017%

references to names or titles, particularly within a specific context

No Comments

No Known Activations