INDEX

Explanations

verbs followed by phrases indicating alternative actions or choices

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Know

-0.64

 acron

-0.59

abre

-0.59

 asses

-0.59

Can

-0.55

 unbeat

-0.54

Got

-0.54

 touches

-0.52

paper

-0.51

 Definitions

-0.50

POSITIVE LOGITS

 instead

2.21

Instead

2.09

 Instead

2.04

instead

1.88

 Rather

1.71

Rather

1.69

rather

1.53

 rather

1.52

 merely

1.35

 simply

1.11

Activations Density 0.996%

verbs followed by phrases indicating alternative actions or choices

No Comments

No Known Activations