INDEX

Explanations

names of Formula One racing drivers

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names of individuals and entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Pow

-0.70

 Action

-0.70

 affiliated

-0.69

iv

-0.69

 affiliate

-0.68

 Native

-0.68

cas

-0.66

 Grizz

-0.66

Lyn

-0.66

 Poly

-0.65

POSITIVE LOGITS

ettel

4.79

 Alonso

1.53

ende

1.37

ussen

1.32

 Ferrari

1.24

 lapt

1.15

 McLaren

1.11

essler

1.10

 Klopp

1.09

enegger

1.09

Activations Density 0.046%

names of Formula One racing drivers

proper nouns, specifically names of individuals and entities

No Comments

No Known Activations