INDEX

Explanations

references to physical health or fitness-related topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

::<

-0.16

okino

-0.15

ibbon

-0.15

adow

-0.15

errer

-0.14

UnitOfWork

-0.14

rup

-0.14

wel

-0.13

ibe

-0.13

firm

-0.13

POSITIVE LOGITS

:///

0.14

 altogether

0.13

 Gerry

0.13

 elev

0.12

Hou

0.12

 scare

0.12

rÅ¾

0.12

/dc

0.12

umber

0.12

ìĤ¬íļĮ

0.12

Activations Density 1.565%

references to physical health or fitness-related topics

No Comments

No Known Activations