INDEX

Explanations

superlatives or comparisons of degree, such as "most" or "more."

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

æ©

-0.90

rompt

-0.88

heid

-0.83

pload

-0.82

 Films

-0.79

oak

-0.78

instead

-0.76

eto

-0.74

undle

-0.74

ategories

-0.74

POSITIVE LOGITS

 important

1.29

 powerful

1.15

 obvious

1.13

likely

1.12

 prominent

1.10

 interesting

1.10

 plausible

1.10

 likely

1.09

 efficient

1.09

 profitable

1.08

Activations Density 10.217%

superlatives or comparisons of degree, such as "most" or "more."

No Comments

No Known Activations