INDEX

Explanations

user interface actions and interactions

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

éĩİ

-0.17

ucus

-0.15

hend

-0.15

wake

-0.14

TEGER

-0.14

wild

-0.14

wed

-0.14

ÑĢÐ¸Ñĩ

-0.14

-Sah

-0.14

beeld

-0.13

POSITIVE LOGITS

 Furn

0.15

pec

0.15

lep

0.14

 Kaynak

0.14

ÐµÑĢÐ¶

0.14

 Lloyd

0.14

ysis

0.14

©

0.13

Ð¾ÑģÐ¿

0.13

Activations Density 0.025%

user interface actions and interactions

No Comments

No Known Activations