INDEX

Explanations

expressions of criticism and evaluation of performance

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

 actually

-0.18

actually

-0.16

æĮ¯ãĤĬ

-0.14

åħ¶å®ŀ

-0.14

leneck

-0.14

 literally

-0.13

lama

-0.13

ennen

-0.13

 colorful

-0.13

aklÄ±

-0.13

POSITIVE LOGITS

igo

0.18

Cra

0.16

plex

0.15

 Broker

0.14

 Bene

0.14

isor

0.14

adolu

0.14

 players

0.14

 complicated

0.14

Datas

0.13

Activations Density 0.013%

expressions of criticism and evaluation of performance

No Comments

No Known Activations