INDEX

Explanations

phrases related to surprise and performance quality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

 corner

-0.14

HLT

-0.14

omed

-0.14

 Rosenstein

-0.14

 Oval

-0.14

exp

-0.13

iggs

-0.13

ÙĦØ§Øª

-0.13

Ð°Ð½ÑĮ

-0.13

ala

-0.13

POSITIVE LOGITS

 socks

0.24

sock

0.18

 SOCK

0.17

ocial

0.17

 Conc

0.16

undler

0.15

park

0.15

 park

0.14

Sock

0.14

convert

0.14

Activations Density 0.039%

phrases related to surprise and performance quality

No Comments

No Known Activations