INDEX

Explanations

verbal mentions of the concept of healing

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of healing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ARGET

-0.78

POL

-0.74

VAL

-0.71

nit

-0.70

VP

-0.67

 GEAR

-0.67

 Stur

-0.65

section

-0.65

Johnson

-0.64

err

-0.64

POSITIVE LOGITS

 heal

1.33

 heals

1.33

 healing

1.29

 healed

1.17

 healer

1.02

 Healing

0.88

 cures

0.87

ilitating

0.86

 wounds

0.86

 Heal

0.82

Activations Density 0.007%

verbal mentions of the concept of healing

terms related to the concept of healing

No Comments

No Known Activations