INDEX

Explanations

references to quantities or rankings of people or things in a list or group

oai_token-act-pair · gpt-3.5-turbo

phrases indicating comparisons or lists involving groups of people

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

section

-0.82

ossession

-0.79

 Simulator

-0.77

lag

-0.75

urgy

-0.74

uria

-0.74

anguage

-0.74

ategory

-0.73

umption

-0.73

obia

-0.71

POSITIVE LOGITS

 celebrities

1.58

 nominees

1.52

 presidents

1.49

 candidates

1.44

 secretaries

1.42

 politicians

1.40

 personalities

1.39

 CEOs

1.37

 prominent

1.32

 predecessors

1.32

Activations Density 0.485%

references to quantities or rankings of people or things in a list or group

phrases indicating comparisons or lists involving groups of people

No Comments

No Known Activations