INDEX

Explanations

symbols indicating questions and emotional expressions

oai_token-act-pair · gpt-3.5-turbo

questions and inquiries throughout the text

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

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

 eleph

-0.80

anwhile

-0.74

 curves

-0.67

 proport

-0.67

 phased

-0.64

 targets

-0.64

ngth

-0.64

sterdam

-0.63

aditional

-0.62

 looph

-0.61

POSITIVE LOGITS

Answer

1.24

Absolutely

1.00

Well

0.96

Probably

0.90

ÂłÂłÂłÂł

0.88

RH

0.86

Honestly

0.85

Yes

0.85

↵

0.84

Yep

0.82

Activations Density 0.110%

symbols indicating questions and emotional expressions

questions and inquiries throughout the text

No Comments

No Known Activations