INDEX

Explanations

references to negative or serious incidents

oai_token-act-pair · gpt-3.5-turbo

references to specific incidents or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ularity

-0.84

emouth

-0.83

azy

-0.80

eday

-0.79

ichick

-0.76

heit

-0.76

rows

-0.75

icrobial

-0.73

tsky

-0.73

veland

-0.72

POSITIVE LOGITS

 involving

1.03

 incident

0.89

 incidents

0.84

 perpetrated

0.79

uality

0.79

uates

0.77

 occurred

0.76

 occurrence

0.73

uation

0.72

UAL

0.71

Activations Density 0.033%

references to negative or serious incidents

references to specific incidents or events

No Comments

No Known Activations