INDEX

Explanations

phrases related to potential outcomes or possibilities

oai_token-act-pair · gpt-3.5-turbo

terms related to expectations, possibilities, and conditions

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

aughs

-0.76

enhagen

-0.70

inho

-0.70

 CrossRef

-0.70

 Braun

-0.69

 Sutherland

-0.69

 Keller

-0.64

 Debor

-0.64

 Charity

-0.64

 Bellev

-0.63

POSITIVE LOGITS

 whatsoever

1.73

 imaginable

1.02

 except

0.88

 besides

0.85

 thereof

0.84

soever

0.81

 semblance

0.78

 pertaining

0.75

 conceivable

0.75

 deemed

0.73

Activations Density 0.198%

phrases related to potential outcomes or possibilities

terms related to expectations, possibilities, and conditions

No Comments

No Known Activations