INDEX

Explanations

terms related to political decisions or actions

oai_token-act-pair · gpt-3.5-turbo

references to policy decisions or legal restrictions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

parts

-0.73

Union

-0.70

ESV

-0.66

avier

-0.66

 Dynamics

-0.65

Met

-0.64

 Rove

-0.64

Fit

-0.63

æĪ¦

-0.63

part

-0.63

POSITIVE LOGITS

 moratorium

1.38

 seiz

1.02

 warr

0.86

 roundup

0.83

isode

0.82

aution

0.81

 destro

0.80

EStream

0.78

raint

0.76

 dismant

0.75

Activations Density 0.010%

terms related to political decisions or actions

references to policy decisions or legal restrictions

No Comments

No Known Activations