INDEX

Explanations

references to sports achievements and teams

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

SEA

-0.17

gaard

-0.16

å¨

-0.15

arella

-0.15

aller

-0.14

ÑĢÐ°Ð²Ð°

-0.14

DBG

-0.14

_lineno

-0.14

 franchise

-0.13

åĦĢ

-0.13

POSITIVE LOGITS

 Olympic

0.27

 gold

0.24

USA

0.23

 Pool

0.21

USA

0.20

 pool

0.20

usa

0.20

 Team

0.19

 Trials

0.19

 Worlds

0.18

Activations Density 0.032%

references to sports achievements and teams

No Comments

No Known Activations