INDEX

Explanations

mentions of built-in features or components within devices or systems

oai_token-act-pair · gpt-3.5-turbo

phrases related to built-in features or functionalities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Mub

-0.92

channelAvailability

-0.74

 Bought

-0.74

aping

-0.71

OPA

-0.71

Ĥª

-0.69

govtrack

-0.69

_-

-0.69

":[

-0.68

 Carbuncle

-0.67

POSITIVE LOGITS

jury

0.77

clusions

0.76

strument

0.70

 alloy

0.70

bodied

0.70

urance

0.65

versely

0.64

 auto

0.64

 custom

0.63

sel

0.63

Activations Density 0.038%

mentions of built-in features or components within devices or systems

phrases related to built-in features or functionalities

No Comments

No Known Activations