INDEX

Explanations

topics related to politics, corruption, space exploration, medical procedures, and waste management

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

,,,,

-0.78

urat

-0.78

ouls

-0.74

aneous

-0.73

olen

-0.73

imov

-0.72

orts

-0.70

ilts

-0.70

aws

-0.69

arez

-0.69

POSITIVE LOGITS

 program

1.23

 programme

1.20

 system

1.15

 portfolio

1.14

 empire

1.12

 regime

1.10

 bureau

1.08

 advisory

1.06

 department

1.06

 industry

1.06

Activations Density 2.036%

topics related to politics, corruption, space exploration, medical procedures, and waste management

No Comments

No Known Activations