INDEX

Explanations

mentions of political figures and controversies, particularly related to Hillary Clinton and government investigations

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

teenth

-0.76

ener

-0.68

 Gareth

-0.68

alam

-0.67

DIS

-0.66

umar

-0.66

ribune

-0.65

semble

-0.65

obe

-0.64

raints

-0.64

POSITIVE LOGITS

 Clinton

1.18

 Rodham

1.14

Clinton

1.13

INTON

1.06

clinton

1.01

 Hillary

0.91

 Supporters

0.88

 impeachment

0.85

 Campaign

0.83

ite

0.82

Activations Density 0.464%

mentions of political figures and controversies, particularly related to Hillary Clinton and government investigations

No Comments

No Known Activations