INDEX

Explanations

words related to raising questions or uncertainties

oai_token-act-pair · gpt-3.5-turbo

references to questions or inquiries raised in a context

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ufact

-0.94

rites

-0.81

urses

-0.77

yss

-0.74

orpor

-0.72

alty

-0.71

tsky

-0.69

emetery

-0.69

assad

-0.68

odka

-0.67

POSITIVE LOGITS

naires

1.29

 unanswered

1.12

naire

0.99

Ans

0.89

 arises

0.88

 questions

0.88

 mark

0.87

 arise

0.86

 posed

0.80

 plag

0.79

Activations Density 0.038%

words related to raising questions or uncertainties

references to questions or inquiries raised in a context

No Comments

No Known Activations