INDEX

Explanations

specific character sequences in text, particularly unusual or accented letters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atabase

-0.17

ãĤ©

-0.16

aut

-0.16

issa

-0.16

ial

-0.15

igh

-0.15

autor

-0.14

mie

-0.14

matrix

-0.14

POSITIVE LOGITS

0.23

ãĥ§

0.18

ãĥ¥

0.17

iado

0.17

 dá»¥

0.16

ecta

0.16

keit

0.16

rvine

0.16

Ð¸

0.15

breaking

0.15

Activations Density 0.085%

specific character sequences in text, particularly unusual or accented letters

No Comments

No Known Activations