INDEX

Explanations

symbols and text formatting

oai_token-act-pair · gpt-3.5-turbo

phrases or terms related to negative evaluations or criticisms

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

'-' token

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

 Dickinson

-0.68

 standing

-0.67

 decon

-0.67

 recourse

-0.65

 manif

-0.65

 respons

-0.64

 beforehand

-0.63

 toget

-0.63

 clen

-0.63

pen

-0.61

POSITIVE LOGITS

 Advertisement

1.23

webkit

1.06

=-=-=-=-=-=-=-=-

1.01

_-

0.94

 Stuff

0.89

 Expand

0.86

Updated

0.85

 Updated

0.85

notation

0.85

=-=-=-=-

0.84

Activations Density 0.023%

symbols and text formatting

phrases or terms related to negative evaluations or criticisms

'-' token

No Comments

No Known Activations