INDEX

Explanations

first person plural pronouns followed by verbs indicating a scholarly or investigative action

oai_token-act-pair · gpt-3.5-turbo

instances of the word "we" and related phrases indicating collective conclusions or observations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Canter

-0.73

 Celebration

-0.72

 Vanity

-0.72

 Crest

-0.70

 dues

-0.69

 Reneg

-0.68

 extortion

-0.67

 Eternal

-0.67

 Guardians

-0.67

 Siege

-0.66

POSITIVE LOGITS

 hypothes

1.59

 hypothesized

1.44

 observe

1.35

 conclude

1.29

 observed

1.28

 speculate

1.28

 analysed

1.23

 analyse

1.18

 studied

1.14

 analyzed

1.12

Activations Density 0.155%

first person plural pronouns followed by verbs indicating a scholarly or investigative action

instances of the word "we" and related phrases indicating collective conclusions or observations

No Comments

No Known Activations