INDEX

Explanations

references to studios and creative workspaces

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

AKE

-0.16

orque

-0.16

udi

-0.16

IU

-0.15

uela

-0.15

ned

-0.14

 ...↵↵↵↵

-0.14

ocl

-0.14

erland

-0.14

lez

-0.14

POSITIVE LOGITS

elling

0.17

ONTAL

0.17

bery

0.16

lain

0.16

engl

0.16

OKIE

0.16

Ø®Ø§ÙĨÙĩ

0.15

-san

0.15

prise

0.15

shade

0.13

Activations Density 0.032%

references to studios and creative workspaces

No Comments

No Known Activations

references to studios and creative workspaces

No Comments

No Known Activations