INDEX

Explanations

technical code snippets with commands or instructions starting with '*' and generally in relation to computing or programming

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

eness

-0.98

adelphia

-0.91

kson

-0.82

ilitarian

-0.82

omorphic

-0.82

ichick

-0.81

 delinqu

-0.79

inelli

-0.79

 boards

-0.79

nesses

-0.78

POSITIVE LOGITS

âĢ¢âĢ¢

1.11

AUT

1.11

SK

1.06

ERROR

1.04

blank

0.99

NEW

0.97

insert

0.96

laughs

0.96

Insert

0.94

DATA

0.92

Activations Density 6.876%

technical code snippets with commands or instructions starting with '*' and generally in relation to computing or programming

No Comments

No Known Activations