INDEX

Explanations

phrases related to past events or behavior patterns

oai_token-act-pair · gpt-3.5-turbo

references to history and track records of entities or individuals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ishable

-0.72

ando

-0.69

wagen

-0.67

oner

-0.66

uri

-0.66

idden

-0.65

asus

-0.63

itely

-0.63

agraph

-0.61

ower

-0.61

POSITIVE LOGITS

 revolving

0.76

 dating

0.70

 spanning

0.69

breaking

0.69

 fraught

0.68

cles

0.68

 favorable

0.67

 stretching

0.67

 brewing

0.67

acqu

0.67

Activations Density 0.081%

phrases related to past events or behavior patterns

references to history and track records of entities or individuals

No Comments

No Known Activations