INDEX

Explanations

mentions of legal or ethical issues

oai_token-act-pair · gpt-3.5-turbo

references to allegations and controversies related to sexual misconduct

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 yourselves

-0.83

respective

-0.75

whatever

-0.69

%:

-0.69

prisingly

-0.69

ardless

-0.68

'/

-0.67

achus

-0.66

inav

-0.66

========

-0.66

POSITIVE LOGITS

 Hurricane

0.82

his

0.81

 hers

0.76

 Proposition

0.74

 chemotherapy

0.74

 inmate

0.72

her

0.71

him

0.70

 contraceptives

0.69

 AIDS

0.67

Activations Density 0.801%

mentions of legal or ethical issues

references to allegations and controversies related to sexual misconduct

No Comments

No Known Activations