INDEX

Explanations

coding or programming terminology related to functions and parameters

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

Ð¼Ð¸ÑĤ

-0.15

ancock

-0.14

queda

-0.14

olem

-0.14

mland

-0.14

 ÎłÎ±Î½

-0.14

modifiable

-0.13

 Culture

-0.13

trand

-0.13

eson

-0.13

POSITIVE LOGITS

 reuse

0.22

reuse

0.21

ibo

0.18

_REUSE

0.18

 secondary

0.18

 already

0.18

Reuse

0.17

 merely

0.16

already

0.16

ientes

0.15

Activations Density 0.154%

coding or programming terminology related to functions and parameters

No Comments

No Known Activations