INDEX

Explanations

terms and phrases associated with high levels or significant quantities

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

rypted

-0.15

ensa

-0.14

AGED

-0.14

 footing

-0.14

-0.13

edList

-0.13

 Trou

-0.13

itler

-0.13

 packet

-0.13

POSITIVE LOGITS

etsk

0.18

quil

0.15

omatic

0.15

sett

0.15

riba

0.15

atus

0.15

.RunWith

0.14

($('<

0.14

orgia

0.14

ocop

0.14

Activations Density 0.013%

terms and phrases associated with high levels or significant quantities

No Comments

No Known Activations