INDEX

Explanations

discussions or passages related to philosophy and ethics

oai_token-act-pair · gpt-3.5-turbo

instances of the word "as" particularly when it precedes the word "to", indicating a purposive or explanatory context

oai_token-act-pair · gpt-3.5-turbo Triggered by @danbraun

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

 millenn

-0.67

ships

-0.63

oller

-0.60

kefeller

-0.58

 Cards

-0.57

osc

-0.57

CCC

-0.55

hospital

-0.54

complex

-0.54

 corrid

-0.53

POSITIVE LOGITS

pects

0.81

 soon

0.74

bestos

0.74

ylum

0.74

ynchron

0.72

 evidenced

0.71

 regards

0.70

pires

0.69

 opposed

0.68

ocial

0.66

Activations Density 7.781%

discussions or passages related to philosophy and ethics

instances of the word "as" particularly when it precedes the word "to", indicating a purposive or explanatory context

No Comments

No Known Activations