INDEX

Explanations

references to specific cultural or entertainment works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

ifest

-0.15

gesch

-0.15

entiful

-0.15

ascript

-0.14

azard

-0.14

entrant

-0.14

almart

-0.14

 upbeat

-0.14

 symbol

-0.14

 cÃ©lib

-0.13

POSITIVE LOGITS

 ÑĢÐ°Ð²Ð½Ð¾

0.17

iram

0.16

 master

0.15

 rival

0.15

 ******************************************************************************↵

0.15

/master

0.15

igli

0.14

ocha

0.14

 equal

0.14

etter

0.14

Activations Density 0.013%

references to specific cultural or entertainment works

No Comments

No Known Activations