INDEX

Explanations

keywords related to news entities and topics

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Autonomous

-0.17

oot

-0.16

ista

-0.14

akes

-0.14

erp

-0.14

vox

-0.14

ë¸Į

-0.14

ouser

-0.13

irit

-0.13

.Z

-0.13

POSITIVE LOGITS

arshal

0.17

imir

0.17

 Corps

0.16

(æ°´

0.15

agram

0.15

ampoline

0.14

abase

0.14

pgsql

0.14

agi

0.14

amage

0.14

Activations Density 0.065%

keywords related to news entities and topics

No Known Activations