INDEX

Explanations

phrases related to destruction or attacks on infrastructure

oai_token-act-pair · gpt-3.5-turbo

measurements and quantities related to donations and resources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Evangel

-0.64

 presiding

-0.61

 Citizen

-0.59

 Colo

-0.59

 Rural

-0.53

 Citizens

-0.53

 feudal

-0.52

 Educ

-0.52

 UNESCO

-0.51

 Syri

-0.51

POSITIVE LOGITS

 increments

0.81

 totaling

0.78

 apiece

0.77

abytes

0.74

inately

0.71

 averaged

0.70

 evenly

0.70

 perpend

0.69

 totals

0.69

 averaging

0.68

Activations Density 1.020%

phrases related to destruction or attacks on infrastructure

measurements and quantities related to donations and resources

No Comments

No Known Activations