INDEX

Explanations

numerical data related to cases and performance statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ohen

-0.15

 Shir

-0.14

urban

-0.14

oven

-0.14

ormal

-0.13

à¸Ńà¸Ł

-0.13

_BT

-0.13

uite

-0.13

jdk

-0.13

quant

-0.13

POSITIVE LOGITS

 total

0.24

total

0.20

 TOTAL

0.19

 already

0.17

_total

0.16

 toplam

0.16

TOTAL

0.16

now

0.16

already

0.16

 Total

0.15

Activations Density 0.213%

numerical data related to cases and performance statistics

No Comments

No Known Activations