INDEX

Explanations

proper nouns related to locations, specifically locations named "Cape"

oai_token-act-pair · gpt-3.5-turbo

mentions of "Cape" related to various geographical locations and contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

omen

-0.84

arial

-0.76

orno

-0.74

amine

-0.73

enegger

-0.73

agi

-0.71

selves

-0.68

giving

-0.67

ignment

-0.67

geist

-0.66

POSITIVE LOGITS

 Cape

1.14

Cod

0.96

 Canaveral

0.87

lla

0.84

 Coral

0.84

 Coast

0.81

Nam

0.80

llan

0.79

 Niger

0.74

 Shoals

0.71

Activations Density 0.009%

proper nouns related to locations, specifically locations named "Cape"

mentions of "Cape" related to various geographical locations and contexts

No Comments

No Known Activations