INDEX

Explanations

numerical sequences associated with steps or instructions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Mun

-0.82

 Daniels

-0.82

 Nashville

-0.81

Ner

-0.80

 Harmon

-0.78

NAS

-0.78

 Gael

-0.76

Gru

-0.74

 Gore

-0.74

ãĥķ

-0.72

POSITIVE LOGITS

 Step

1.51

 STEP

1.49

Step

1.45

step

1.32

STEP

1.29

 Stephens

1.28

 Steps

1.21

 Stage

1.20

Pt

1.18

 step

1.16

Activations Density 0.224%

numerical sequences associated with steps or instructions

No Comments

No Known Activations