INDEX

Explanations

terms related to accountability, such as "accountability" itself and phrases mentioning the need for or instances of accountability

oai_token-act-pair · gpt-3.5-turbo

mentions of accountability in various contexts

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

ker

-0.76

enegger

-0.73

othe

-0.69

eds

-0.69

ken

-0.68

ke

-0.68

nee

-0.68

yon

-0.68

othes

-0.67

eding

-0.67

POSITIVE LOGITS

 accountability

1.07

 Accountability

0.93

rity

0.83

parency

0.83

ilogy

0.81

 destro

0.80

eatures

0.79

acies

0.79

¥ŀ

0.79

 sqor

0.79

Activations Density 0.005%

terms related to accountability, such as "accountability" itself and phrases mentioning the need for or instances of accountability

mentions of accountability in various contexts

No Comments

No Known Activations