INDEX

Explanations

uses of quantifiers like 'few', 'many', and 'some' in a political context

oai_token-act-pair · gpt-3.5-turbo

referential terms that denote collective groups and their attributes

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

regnancy

-0.91

ategory

-0.84

andise

-0.84

ossession

-0.81

igation

-0.81

ogenesis

-0.76

ãĤ´ãĥ³

-0.76

otation

-0.75

atform

-0.75

yrinth

-0.74

POSITIVE LOGITS

 idiots

1.48

 guys

1.47

 dudes

1.44

 folks

1.42

 gentlemen

1.35

 fellows

1.33

 losers

1.31

 people

1.27

 individuals

1.27

 fools

1.26

Activations Density 0.395%

uses of quantifiers like 'few', 'many', and 'some' in a political context

referential terms that denote collective groups and their attributes

No Comments

No Known Activations