INDEX

Explanations

names and titles in various contexts, particularly in relation to arts and media

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

oined

-0.17

ieu

-0.15

fv

-0.14

akra

-0.14

emale

-0.14

 localVar

-0.14

ono

-0.14

ugar

-0.14

arkin

-0.14

POSITIVE LOGITS

amps

0.17

ê²

0.14

amp

0.14

ãģªãģĮãĤī

0.14

 pres

0.14

appen

0.13

ÐµÐ·

0.13

_fname

0.13

.isActive

0.13

ppo

0.13

Activations Density 0.036%

names and titles in various contexts, particularly in relation to arts and media

No Comments

No Known Activations