INDEX

Explanations

mentions of political entities and events, such as political statements and election results

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

lethal

-0.83

 certs

-0.75

Prosecut

-0.73

 restructuring

-0.72

 reorgan

-0.68

washer

-0.68

REF

-0.66

 multiplication

-0.65

 chops

-0.65

NSA

-0.64

POSITIVE LOGITS

 Alexandria

0.95

 Bethlehem

0.85

 southwestern

0.85

 northeastern

0.84

 Toronto

0.84

 northern

0.83

 southern

0.83

 Plymouth

0.83

 Northeast

0.82

 Crimea

0.81

Activations Density 0.084%

mentions of political entities and events, such as political statements and election results

No Comments

No Known Activations