INDEX

Explanations

locations and buildings

oai_token-act-pair · gpt-3.5-turbo

references to location or place-related details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ĻĤ

-0.85

£ı

-0.71

DSM

-0.69

 Gamergate

-0.68

Pax

-0.68

 Solitaire

-0.67

 Turing

-0.67

 Guant

-0.67

 Rowling

-0.67

 Dresden

-0.67

POSITIVE LOGITS

 tents

0.85

unicip

0.76

taboola

0.73

 floats

0.73

urized

0.73

ulously

0.72

 barric

0.72

 tourists

0.71

ominium

0.70

 crews

0.69

Activations Density 1.007%

locations and buildings

references to location or place-related details

No Comments

No Known Activations