INDEX

Explanations

titles of scientific papers

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

ulhu

-0.72

ufact

-0.62

 Bulgar

-0.61

rity

-0.60

Zin

-0.59

FACE

-0.59

 Gillespie

-0.57

eele

-0.56

eus

-0.55

loaded

-0.55

POSITIVE LOGITS

comfort

0.59

fort

0.58

jad

0.57

ression

0.57

quest

0.57

 hesitate

0.56

 slightest

0.54

 aback

0.54

content

0.53

irm

0.53

Activations Density 6.820%

titles of scientific papers

No Comments

No Known Activations