INDEX

Explanations

words related to the idea of creating or taking action towards improvement or achievement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

indir

-0.15

.Restrict

-0.14

anuts

-0.14

ientes

-0.14

odings

-0.14

icit

-0.14

AMI

-0.14

hangi

-0.13

inan

-0.13

ags

-0.13

POSITIVE LOGITS

 sure

0.40

sure

0.31

 Sure

0.30

Sure

0.27

 connections

0.22

SURE

0.20

 dreams

0.19

Connections

0.19

 Connections

0.18

 waves

0.18

Activations Density 0.076%

words related to the idea of creating or taking action towards improvement or achievement

No Comments

No Known Activations