INDEX

Explanations

terms related to race-related metrics and activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

olik

-0.18

ÐºÑĥÐ»

-0.16

Ø¨Ø§ØŃ

-0.16

igner

-0.16

uries

-0.16

linkplain

-0.16

ieÅ¼

-0.15

alat

-0.14

hk

-0.14

POSITIVE LOGITS

lap

0.32

pit

0.31

Lap

0.29

lap

0.28

 pits

0.27

Pit

0.27

pit

0.24

 laps

0.24

 strategy

0.21

 caution

0.21

Activations Density 0.020%

terms related to race-related metrics and activities

No Comments

No Known Activations