INDEX

Explanations

names of locations and events

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

thood

-0.85

natureconservancy

-0.73

ertodd

-0.71

obin

-0.67

xual

-0.66

versive

-0.66

osexual

-0.66

rez

-0.65

reated

-0.65

onym

-0.65

POSITIVE LOGITS

 Calif

1.23

 Colo

1.12

TX

1.02

Fla

1.00

TN

0.99

Md

0.98

Ala

0.98

 Tenn

0.97

Ont

0.96

 Illinois

0.94

Activations Density 0.574%

names of locations and events

No Comments

No Known Activations