INDEX

Explanations

references to physical scars

oai_token-act-pair · gpt-3.5-turbo

references to scars, both physical and metaphorical

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Unch

-0.64

eering

-0.63

eers

-0.61

 Bulldogs

-0.59

ysis

-0.59

 stabilization

-0.57

arantine

-0.57

 deliveries

-0.57

 desperate

-0.57

 Standards

-0.57

POSITIVE LOGITS

red

1.27

crow

1.08

lett

1.07

ring

1.07

ves

1.06

fing

1.06

fed

1.02

face

1.02

lets

1.01

nton

1.01

Activations Density 0.032%

references to physical scars

references to scars, both physical and metaphorical

No Comments

No Known Activations