INDEX

Explanations

phrases related to turning something on or off

oai_token-act-pair · gpt-3.5-turbo

commands related to enabling or disabling functionality

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

enegger

-0.81

 Shards

-0.75

oru

-0.71

Transaction

-0.71

çĶ°

-0.70

istani

-0.68

BTC

-0.67

loe

-0.66

ORTS

-0.66

esan

-0.65

POSITIVE LOGITS

 automatic

0.83

 lights

0.81

screen

0.79

 alarms

0.78

 autop

0.78

 mute

0.78

 ignition

0.76

 Genie

0.76

 blinking

0.74

 alarm

0.74

Activations Density 0.057%

phrases related to turning something on or off

commands related to enabling or disabling functionality

No Comments

No Known Activations