INDEX

Explanations

references to size, scale, volume, and numbers

oai_token-act-pair · gpt-3.5-turbo

references to the size or scale of objects, issues, or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

selves

-0.69

 Ending

-0.67

 Scenes

-0.61

 Anthem

-0.59

 FANTASY

-0.59

onis

-0.59

lynn

-0.58

 Genie

-0.58

icus

-0.57

Progress

-0.57

POSITIVE LOGITS

 nature

1.13

 cumbers

1.08

 inherent

0.94

 similarities

0.93

itiveness

0.93

availability

0.90

ability

0.89

 size

0.88

 similarity

0.86

lessness

0.86

Activations Density 0.367%

references to size, scale, volume, and numbers

references to the size or scale of objects, issues, or situations

No Comments

No Known Activations