INDEX

Explanations

negative declarations or negations

oai_token-act-pair · gpt-3.5-turbo

repeated expressions of negation or the absence of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

haul

-0.80

antis

-0.70

UME

-0.69

Broad

-0.69

Stra

-0.68

asts

-0.67

rib

-0.66

orts

-0.65

CD

-0.65

ordes

-0.65

POSITIVE LOGITS

 appetite

0.81

 shortage

0.79

 overlap

0.77

 precedent

0.76

 hesitation

0.74

 unanim

0.73

 evidence

0.73

 chance

0.72

 detectable

0.71

 buried

0.67

Activations Density 0.086%

negative declarations or negations

repeated expressions of negation or the absence of something

No Comments

No Known Activations