INDEX

Explanations

phrases related to interactions or events involving neighbors

oai_token-act-pair · gpt-3.5-turbo

dialogue or quotes indicating communication between individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Airbus

-0.82

apter

-0.81

 Streaming

-0.79

 onstage

-0.79

pter

-0.77

pod

-0.77

selage

-0.77

 Titanic

-0.77

 Trance

-0.76

TIT

-0.75

POSITIVE LOGITS

 neighbors

1.80

 residents

1.76

Residents

1.76

Neigh

1.75

 Residents

1.71

 neighborhood

1.68

 township

1.64

 Neigh

1.61

 Neighborhood

1.55

 neighbor

1.54

Activations Density 0.612%

phrases related to interactions or events involving neighbors

dialogue or quotes indicating communication between individuals

No Comments

No Known Activations