INDEX

Explanations

references to parody and humor related to entertainment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

 Trim

-0.16

avou

-0.15

TL

-0.15

agas

-0.14

anten

-0.14

ensburg

-0.14

Forever

-0.14

agate

-0.14

Ris

-0.14

andan

-0.14

POSITIVE LOGITS

0.15

(~(

0.15

ysi

0.15

aur

0.15

epar

0.15

0.14

enga

0.14

ock

0.14

ossa

0.14

 Transparent

0.14

Activations Density 0.311%

references to parody and humor related to entertainment

No Comments

No Known Activations