INDEX

Explanations

negative and judgmental statements

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "is" followed by various attributes or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ffe

-0.77

ventures

-0.76

 grips

-0.72

 Bunker

-0.70

ãĤ´ãĥ³

-0.70

atl

-0.69

Scroll

-0.69

 Tune

-0.67

pled

-0.67

 Hitch

-0.65

POSITIVE LOGITS

 indistinguishable

1.31

 considered

1.23

 liable

1.20

 acceptable

1.18

 preferable

1.14

 unworthy

1.13

 inherently

1.13

 invalid

1.10

 unacceptable

1.10

 deemed

1.09

Activations Density 0.290%

negative and judgmental statements

instances of the phrase "is" followed by various attributes or conditions

No Comments

No Known Activations