INDEX

Explanations

specific named entities or titles

oai_token-act-pair · gpt-3.5-turbo

references to specific applications or programs, particularly in technology and media contexts

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

 wealthy

-0.84

 compan

-0.83

 notor

-0.79

 customary

-0.78

 examiner

-0.76

 inclined

-0.74

 manufacturer

-0.74

akespe

-0.74

 clerks

-0.71

 subdiv

-0.71

POSITIVE LOGITS

 Presents

1.32

 Reloaded

1.15

 Dreams

1.10

 Skies

1.08

 Adventures

1.07

 Basics

1.07

 Rescue

1.07

 Awareness

1.06

 Strategies

1.06

 Fever

1.05

Activations Density 0.492%

specific named entities or titles

references to specific applications or programs, particularly in technology and media contexts

No Comments

No Known Activations