INDEX

Explanations

numerical statistics or data points related to performance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

-0.15

buz

-0.15

alu

-0.14

atcher

-0.14

agnet

-0.14

 central

-0.13

á»¥c

-0.13

 tilt

-0.13

verb

-0.13

odu

-0.13

POSITIVE LOGITS

Loan

0.16

å£²

0.15

ÑĥÐ¹

0.14

utter

0.14

 Loan

0.14

 mouths

0.14

å¹¹

0.14

InBackground

0.14

ationale

0.14

 loan

0.14

Activations Density 0.007%

numerical statistics or data points related to performance

No Comments

No Known Activations