INDEX

Explanations

phrases related to list or description of items or details

oai_token-act-pair · gpt-3.5-turbo

components of articles or reports that include specific details or assertions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Sport

-0.79

ickr

-0.75

assetsadobe

-0.71

Shop

-0.67

VM

-0.64

imeo

-0.63

INESS

-0.62

sis

-0.62

fm

-0.62

vc

-0.61

POSITIVE LOGITS

 elements

0.80

 spoilers

0.76

 redundant

0.76

ographic

0.75

 remnants

0.73

 annotations

0.71

rawdownloadcloneembedreportprint

0.70

 clauses

0.70

ographical

0.70

 prominently

0.70

Activations Density 0.350%

phrases related to list or description of items or details

components of articles or reports that include specific details or assertions

No Comments

No Known Activations