INDEX

Explanations

references to anonymity or anonymous individuals

oai_token-act-pair · gpt-3.5-turbo

references to anonymity 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 @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

hold

-0.70

 Landing

-0.69

Frames

-0.69

akings

-0.65

 Land

-0.65

 Combo

-0.65

rises

-0.65

 grips

-0.64

 gear

-0.63

rients

-0.62

POSITIVE LOGITS

 anonymous

3.79

 anonymously

2.48

 anonymity

2.27

 unnamed

1.99

Anonymous

1.90

 Anonymous

1.84

onymous

1.83

 anonym

1.81

 unidentified

1.58

 confidential

1.49

Activations Density 0.024%

references to anonymity or anonymous individuals

references to anonymity in various contexts

No Known Activations