INDEX

Explanations

words related to governmental policies and legal matters

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

edIn

-0.70

Jaw

-0.68

 unborn

-0.68

 Lauder

-0.67

 Retrieved

-0.66

ORED

-0.64

GGGG

-0.64

rot

-0.63

ieri

-0.63

ered

-0.63

POSITIVE LOGITS

visor

1.41

imposed

1.33

nova

1.26

visors

1.23

marine

1.09

markets

1.08

cell

1.07

massive

1.04

charged

1.04

charg

1.03

Activations Density 1.503%

words related to governmental policies and legal matters

No Comments

No Known Activations