INDEX

Explanations

references to soldiers and military-related content

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

TERN

-0.78

omial

-0.72

âĶĢâĶĢâĶĢâĶĢ

-0.70

aminer

-0.66

CAST

-0.63

tera

-0.62

 srfAttach

-0.60

Solution

-0.60

âĶģ

-0.58

SOURCE

-0.57

POSITIVE LOGITS

 stationed

0.97

 fatig

0.85

 uniforms

0.73

 barracks

0.73

 garrison

0.72

doms

0.66

 guarding

0.66

 soldiers

0.66

 reserv

0.65

 colonel

0.64

Activations Density 7.825%

references to soldiers and military-related content

No Comments

No Known Activations