INDEX

Explanations

ambiguous terms

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

©¶æ

-0.74

 guiActiveUn

-0.73

 Onion

-0.72

pse

-0.70

THR

-0.68

TERN

-0.67

 numbering

-0.67

 Farrell

-0.67

Cow

-0.67

 Ples

-0.65

POSITIVE LOGITS

ished

0.82

ulatory

0.81

uously

0.81

ascus

0.78

olate

0.77

ufact

0.76

ante

0.75

ishment

0.74

illions

0.74

arine

0.72

Activations Density 0.065%

ambiguous terms

No Comments

No Known Activations