INDEX

Explanations

references to sports events or athletes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

-ball

-0.18

 ball

-0.17

 Toronto

-0.16

Toronto

-0.15

Îŀ

-0.15

xes

-0.15

Kan

-0.14

 ì¶ľìŀ¥

-0.14

Kaw

-0.14

Ð¾Ñī

-0.14

POSITIVE LOGITS

ICE

0.19

ICE

0.17

 Blackhawks

0.17

ice

0.16

ice

0.16

anford

0.15

inkel

0.15

 skate

0.15

 Tigers

0.15

Ice

0.15

Activations Density 0.041%

references to sports events or athletes

No Comments

No Known Activations