INDEX

Explanations

keywords related to legal terms and political figures

oai_token-act-pair · gpt-3.5-turbo

phrases related to combating gender bias and promoting equality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 vulner

-0.60

 rigging

-0.60

 ..."

-0.59

 respectively

-0.59

 prec

-0.56

 fame

-0.55

 opposite

-0.53

 polar

-0.51

 persuasion

-0.49

 decisive

-0.49

POSITIVE LOGITS

ashtra

0.70

arius

0.68

arij

0.66

yna

0.65

yn

0.64

 Profile

0.64

abus

0.63

ulous

0.62

azel

0.62

ibliography

0.61

Activations Density 1.642%

keywords related to legal terms and political figures

phrases related to combating gender bias and promoting equality

No Comments

No Known Activations