INDEX

Explanations

phrases related to experiences or experiments

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Cth

-0.70

 Patriarch

-0.67

Nun

-0.64

 virtue

-0.64

 Paste

-0.63

 dwar

-0.63

 Shack

-0.62

 cooker

-0.61

 clad

-0.61

 Skydragon

-0.60

POSITIVE LOGITS

ienced

1.88

iments

1.62

iment

1.58

iences

1.55

ience

1.53

imental

1.44

ts

1.25

ient

1.25

ients

1.10

ien

1.10

Activations Density 0.035%

phrases related to experiences or experiments

No Comments

No Known Activations