INDEX

Explanations

bibliographic or publication-related information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

 scheme

-0.15

.locals

-0.15

ãĥ³ãĤ¯

-0.14

bett

-0.14

 Scheme

-0.14

innie

-0.14

ä»ĭ

-0.13

 Amber

-0.13

ysa

-0.13

omy

-0.13

POSITIVE LOGITS

èĳĹ

0.17

loth

0.16

_codegen

0.15

utow

0.14

ÃĸL

0.14

EMPL

0.14

å¯¸

0.14

itle

0.14

ãģĿ

0.13

 Spell

0.13

Activations Density 0.025%

bibliographic or publication-related information

No Comments

No Known Activations