INDEX

Explanations

words related to conditions, whether they are conditional or unconditional

oai_token-act-pair · gpt-3.5-turbo

linguistic constructs related to conditions and stipulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

geist

-0.87

natureconservancy

-0.76

 Thro

-0.76

gun

-0.75

 Illustrated

-0.75

conom

-0.75

aos

-0.70

ICH

-0.70

azines

-0.69

Club

-0.69

POSITIVE LOGITS

 confir

1.10

etheless

0.97

 conditional

0.92

 unconditional

0.86

 uncond

0.85

rontal

0.78

 probabilities

0.76

 lamb

0.75

 sealing

0.75

 pard

0.74

Activations Density 0.018%

words related to conditions, whether they are conditional or unconditional

linguistic constructs related to conditions and stipulations

No Comments

No Known Activations