INDEX

Explanations

mentions of governmental departments or agencies, particularly focusing on legal or investigative contexts

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 worsh

-0.84

 brill

-0.75

 descent

-0.73

 descendant

-0.72

 descend

-0.70

 convent

-0.67

 paradise

-0.67

 neighbourhood

-0.66

 neighbourhoods

-0.65

 coffin

-0.64

POSITIVE LOGITS

 Secretary

0.79

 Accountability

0.78

Ĳ

0.78

 Inspect

0.78

 Commerce

0.76

 Bureau

0.76

Of

0.75

uthor

0.74

artments

0.73

 Inspector

0.73

Activations Density 7.221%

mentions of governmental departments or agencies, particularly focusing on legal or investigative contexts

No Comments

No Known Activations