INDEX

Explanations

discussions about preferences and qualities of various items or experiences

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

ãĥĲãĤ¤

-0.16

rowsable

-0.15

gli

-0.15

iÃª

-0.15

aeper

-0.14

ukkan

-0.14

GuidId

-0.14

 disappoint

-0.14

ivec

-0.14

amaz

-0.14

POSITIVE LOGITS

fit

0.25

 suits

0.25

 lent

0.24

 lend

0.24

 appeals

0.24

 appealed

0.24

 fits

0.23

 lending

0.22

 appeal

0.22

 works

0.21

Activations Density 0.202%

discussions about preferences and qualities of various items or experiences

No Comments

No Known Activations