INDEX

Explanations

proper nouns and names, potentially related to political or historical figures

oai_token-act-pair · gpt-3.5-turbo

references to notable political or historical figures and events

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

 depends

-0.58

 NETWORK

-0.58

:(

-0.53

iquette

-0.52

 polarized

-0.51

 Sloan

-0.50

FO

-0.50

 Freeze

-0.49

ipeg

-0.49

 Decay

-0.48

POSITIVE LOGITS

soDeliveryDate

0.78

pron

0.76

 Leilan

0.72

çĶ°

0.70

ãĥł

0.69

anu

0.64

 deceased

0.63

[|

0.63

Ò

0.59

à¦

0.58

Activations Density 1.633%

proper nouns and names, potentially related to political or historical figures

references to notable political or historical figures and events

No Comments

No Known Activations