INDEX

Explanations

variations of the word "complete" in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ra

-0.19

rias

-0.18

rus

-0.17

riet

-0.17

ski

-0.16

ãĥ³ãĤº

-0.16

sn

-0.15

ses

-0.15

chia

-0.15

 commute

-0.15

POSITIVE LOGITS

ments

0.28

ting

0.22

tes

0.22

ter

0.21

tableFuture

0.21

tem

0.19

ct

0.19

tement

0.19

mente

0.18

ple

0.18

Activations Density 0.006%

variations of the word "complete" in different contexts

No Comments

No Known Activations