INDEX

Explanations

references to political or governmental entities and their relationships

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

veyard

-0.93

okin

-0.82

potion

-0.81

erb

-0.78

owler

-0.77

renheit

-0.77

 Fountain

-0.76

kson

-0.75

Wheel

-0.73

elight

-0.72

POSITIVE LOGITS

 governments

1.01

hips

0.96

 ratified

0.91

 embassies

0.90

 countries

0.88

 nationals

0.87

 nations

0.86

 capitals

0.85

 capit

0.85

 Parties

0.84

Activations Density 0.044%

references to political or governmental entities and their relationships

No Comments

No Known Activations