INDEX

Explanations

terms related to the concept of 'ordinary'

oai_token-act-pair · gpt-3.5-turbo

references to "ordinary" or "seminary" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

GO

-0.69

 shroud

-0.66

cap

-0.65

 Shah

-0.64

 Gear

-0.64

cap

-0.64

 caps

-0.64

 Cthulhu

-0.64

hug

-0.63

 buoy

-0.63

POSITIVE LOGITS

inary

4.83

inarily

2.35

inar

1.35

ordinary

1.31

inous

1.27

inal

1.27

ination

1.22

inances

1.09

ined

1.07

inery

1.03

Activations Density 0.010%

terms related to the concept of 'ordinary'

references to "ordinary" or "seminary" across various contexts

No Comments

No Known Activations