INDEX

Explanations

the word "viable" in contexts where something is considered to be a realistic or workable option

oai_token-act-pair · gpt-3.5-turbo

the term "viable" 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 @ 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

bara

-0.91

yer

-0.82

hun

-0.73

bows

-0.73

Hon

-0.70

alted

-0.70

yang

-0.68

orig

-0.67

andra

-0.67

roy

-0.66

POSITIVE LOGITS

 viable

1.06

 alternatives

0.95

 viability

0.89

GoldMagikarp

0.76

 reuse

0.75

 feas

0.72

 proposition

0.71

 credible

0.69

 alternative

0.68

izable

0.67

Activations Density 0.008%

the word "viable" in contexts where something is considered to be a realistic or workable option

the term "viable" in various contexts

No Comments

No Known Activations