INDEX

Explanations

phrases related to analysis and the evaluation of findings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

FromClass

-0.15

 putas

-0.14

imer

-0.14

.mit

-0.13

çĸĳ

-0.13

acus

-0.13

 nÃ¡v

-0.13

\grid

-0.13

mercial

-0.13

acos

-0.13

POSITIVE LOGITS

 trend

0.16

è±Ĭ

0.15

Ń

0.15

ignon

0.15

Fry

0.14

 Tomorrow

0.13

 Dalton

0.13

ookie

0.13

ÙħØ§ÙĨ

0.13

Activations Density 0.090%

phrases related to analysis and the evaluation of findings

No Comments

No Known Activations