INDEX

Explanations

references to numerical values and levels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

MEA

-0.18

olet

-0.16

xDA

-0.16

DataMember

-0.15

inha

-0.15

Ã©sz

-0.15

bet

-0.14

Î»Î¿Ïį

-0.14

(DWORD

-0.13

 Bilim

-0.13

POSITIVE LOGITS

 admission

0.23

 Admission

0.19

ibold

0.18

 admitted

0.18

 senate

0.17

 Sandwich

0.17

 Batch

0.17

_batches

0.17

 screening

0.16

 cutoff

0.16

Activations Density 0.039%

references to numerical values and levels

No Comments

No Known Activations