INDEX

Explanations

the word "least" followed by a number or a possessive form of the word "least"

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

ses

-0.87

kas

-0.82

 Reborn

-0.82

gnu

-0.79

clerosis

-0.79

halla

-0.79

SHIP

-0.78

xxxxxxxx

-0.78

raltar

-0.78

hots

-0.76

POSITIVE LOGITS

erest

0.94

 amount

0.94

bit

0.91

 practicable

0.91

 intrusive

0.90

 possible

0.89

 conceivable

0.87

 expensive

0.86

 conspicuous

0.86

 forgiving

0.86

Activations Density 7.360%

the word "least" followed by a number or a possessive form of the word "least"

No Comments

No Known Activations