INDEX

Explanations

references to unauthorized disclosures or information breaches

oai_token-act-pair · gpt-3.5-turbo

references to information leaks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iterranean

-0.82

oran

-0.76

stad

-0.67

igel

-0.65

oola

-0.65

aldi

-0.63

ulia

-0.62

ellen

-0.62

lov

-0.61

==

-0.61

POSITIVE LOGITS

Leaks

0.83

 leaks

0.83

 leak

0.77

 leaking

0.76

 leaked

0.75

acious

0.75

ster

0.73

 confidential

0.69

ariat

0.68

 closet

0.67

Activations Density 0.024%

references to unauthorized disclosures or information breaches

references to information leaks

No Comments

No Known Activations