INDEX

Explanations

phrases related to direct interpersonal interactions

oai_token-act-pair · gpt-3.5-turbo

expressions related to direct confrontations or opposing views

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 flares

-0.76

 overc

-0.76

 Reloaded

-0.75

 sooner

-0.74

 Cascade

-0.73

 smugglers

-0.73

 rebuilt

-0.73

 clues

-0.72

 Rohingya

-0.72

 renewed

-0.71

POSITIVE LOGITS

prison

1.45

date

1.44

tail

1.41

surface

1.41

toe

1.40

air

1.38

death

1.37

mouth

1.34

life

1.34

ground

1.33

Activations Density 0.050%

phrases related to direct interpersonal interactions

expressions related to direct confrontations or opposing views

No Comments

No Known Activations