INDEX

Explanations

profane or offensive language

oai_token-act-pair · gpt-3.5-turbo

instances of expletives or strong language

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

irms

-0.69

esthesia

-0.66

eting

-0.66

 breeze

-0.65

onymous

-0.64

 intrusion

-0.64

eca

-0.64

xual

-0.62

 spelling

-0.62

acles

-0.62

POSITIVE LOGITS

**

0.99

***

0.90

ãĥĥãĥī

0.79

cause

0.79

@@@@@@@@

0.74

URI

0.73

adr

0.73

********************************

0.73

****

0.72

keeper

0.72

Activations Density 0.050%

profane or offensive language

instances of expletives or strong language

No Comments

No Known Activations