INDEX

Explanations

words related to political events and actions

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

Zip

-0.72

 retrie

-0.71

 doesnt

-0.70

(?,

-0.69

itably

-0.67

ministic

-0.66

inates

-0.65

iors

-0.65

iol

-0.65

qus

-0.65

POSITIVE LOGITS

 controversies

1.32

 controversy

1.32

 contentious

1.19

 crackdown

1.14

 ongoing

1.08

 controversial

1.05

 scandals

1.05

 recent

1.03

 backlash

1.01

 hotly

1.00

Activations Density 2.461%

words related to political events and actions

No Comments

No Known Activations