INDEX

Explanations

phrases related to the concept of accuracy and correctness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

laz

-0.17

edeki

-0.15

yles

-0.15

ÙĬ

-0.15

thing

-0.15

ATAB

-0.15

Ð¼Ð¾ÑĢ

-0.14

ella

-0.14

íģ

-0.14

marked

-0.14

POSITIVE LOGITS

itude

0.30

 representations

0.23

 representation

0.22

 portrayal

0.22

zza

0.21

itudes

0.21

 depiction

0.19

ives

0.18

Representation

0.17

amente

0.17

Activations Density 0.051%

phrases related to the concept of accuracy and correctness

No Comments

No Known Activations