INDEX

Explanations

phrases related to politics and government

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

 Maced

-0.76

 buoy

-0.75

 Valhalla

-0.74

 Elys

-0.69

orn

-0.67

 clut

-0.67

 flex

-0.66

 rounded

-0.66

 brim

-0.64

 seiz

-0.64

POSITIVE LOGITS

sic

1.67

insert

1.33

REDACTED

1.24

laughs

1.23

being

1.20

blank

1.13

his

1.12

emphasis

1.12

their

1.10

the

1.08

Activations Density 0.581%

phrases related to politics and government

No Comments

No Known Activations