INDEX

Explanations

references to a specific popular franchise in the realm of media and entertainment

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

utics

-0.82

terday

-0.75

LIA

-0.74

essee

-0.73

xual

-0.73

mble

-0.73

éĹĺ

-0.71

ause

-0.71

inventoryQuantity

-0.70

UME

-0.70

POSITIVE LOGITS

 Wars

1.15

light

1.13

buck

1.10

 Trek

1.08

fish

1.07

fighter

1.05

ring

1.05

burst

1.02

Wars

1.00

vation

1.00

Activations Density 0.019%

references to a specific popular franchise in the realm of media and entertainment

No Comments

No Known Activations