INDEX

Explanations

phrases describing the reactions or opinions of fans

oai_token-act-pair · gpt-3.5-turbo

references to fans and their enthusiasm or reactions

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

catentry

-0.83

 treaties

-0.82

 srfAttach

-0.81

 patents

-0.73

 successor

-0.69

è¦ļéĨĴ

-0.68

Ops

-0.67

 exceptions

-0.67

 regulations

-0.67

 extraction

-0.66

POSITIVE LOGITS

 clam

0.88

selves

0.81

 cheering

0.76

 loud

0.74

mire

0.73

raged

0.71

 Bucks

0.70

 Lowry

0.70

facebook

0.68

 incred

0.67

Activations Density 0.664%

phrases describing the reactions or opinions of fans

references to fans and their enthusiasm or reactions

No Comments

No Known Activations