INDEX

Explanations

phrases related to services or service-related activities

oai_token-act-pair · gpt-3.5-turbo

references to various services provided

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

66666666

-0.67

sbm

-0.66

arest

-0.65

ead

-0.64

-0.63

reat

-0.62

Filename

-0.61

 Petr

-0.61

luaj

-0.61

POSITIVE LOGITS

 outage

0.96

 service

0.94

service

0.93

 provider

0.90

 providers

0.90

 Provider

0.88

 Employees

0.81

ibaba

0.79

 Samurai

0.78

 acad

0.78

Activations Density 0.026%

phrases related to services or service-related activities

references to various services provided

No Comments

No Known Activations