INDEX

Explanations

words related to bodily fluids and physical conditions like sweat, tears, and blood

oai_token-act-pair · gpt-3.5-turbo

references to bodily fluids, particularly blood and sweat

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

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

ARM

-0.82

Reviewer

-0.74

 vocational

-0.71

 pivot

-0.69

 geared

-0.67

 demolition

-0.65

onom

-0.64

Solar

-0.64

 feasibility

-0.64

TOR

-0.64

POSITIVE LOGITS

 stains

1.51

 spilled

1.41

 dripping

1.30

 soaked

1.22

 saliva

1.18

 spilling

1.17

stained

1.15

 stain

1.14

 sweat

1.13

plets

1.12

Activations Density 0.245%

words related to bodily fluids and physical conditions like sweat, tears, and blood

references to bodily fluids, particularly blood and sweat

No Comments

No Known Activations