INDEX

Explanations

words related to scientific research or technical documents

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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

rome

-0.93

fo

-0.91

cin

-0.90

ramer

-0.89

BALL

-0.89

ondo

-0.87

oak

-0.87

cutting

-0.87

rip

-0.86

cipled

-0.86

POSITIVE LOGITS

 thereto

1.32

 response

1.08

ivated

1.07

 favorably

1.02

 responses

1.01

iments

1.00

ively

0.98

 affirm

0.96

isson

0.95

naires

0.93

Activations Density 1.587%

words related to scientific research or technical documents

No Comments

No Known Activations