INDEX

Explanations

statistical results presented in a technical and research-oriented context

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

achus

-0.93

anas

-0.84

ourses

-0.84

ataka

-0.82

udeb

-0.81

aneers

-0.80

berman

-0.79

 "$:/

-0.77

assic

-0.77

raltar

-0.76

POSITIVE LOGITS

-.

0.97

 WRITE

0.81

 inhibitor

0.80

ly

0.79

0.78

0.77

0.76

 Braun

0.71

 inhibitors

0.71

Activations Density 11.626%

statistical results presented in a technical and research-oriented context

No Comments

No Known Activations