INDEX

Explanations

phrases related to routines or regular occurrences

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

 exits

-0.80

 skirts

-0.68

 References

-0.64

helm

-0.64

 sins

-0.63

 woes

-0.63

ãĥ©ãĥ³

-0.62

 Commands

-0.62

 warnings

-0.62

 Tears

-0.62

POSITIVE LOGITS

 basis

1.65

 scale

1.24

 whim

1.05

Scale

0.94

 dime

0.94

 hourly

0.94

 continuum

0.92

scale

0.92

cale

0.89

 level

0.87

Activations Density 0.120%

phrases related to routines or regular occurrences

No Comments

No Known Activations