INDEX

Explanations

technical terms and actions related to programming and software behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

tÃ¼

-0.16

opr

-0.16

ouden

-0.15

eniable

-0.15

Zub

-0.14

tir

-0.14

andles

-0.13

ALTH

-0.13

 evenly

-0.13

Tow

-0.13

POSITIVE LOGITS

 directly

0.24

via

0.21

 inside

0.21

 direct

0.21

 manual

0.19

 transparent

0.19

 MANUAL

0.19

 DIRECT

0.18

via

0.18

inside

0.18

Activations Density 0.321%

technical terms and actions related to programming and software behavior

No Comments

No Known Activations