INDEX

Explanations

phrases indicating personal experience or preference in a positive context

oai_token-act-pair · gpt-3.5-turbo

various elements related to storytelling and narrative structure

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

 Ragnar

-0.87

obyl

-0.85

 Scient

-0.82

 Chern

-0.82

 Mald

-0.81

 Rudolph

-0.79

 âĶľâĶĢâĶĢ

-0.78

haul

-0.77

HTTP

-0.76

busters

-0.75

POSITIVE LOGITS

 pens

1.93

pen

1.92

ink

1.84

nib

1.81

Ink

1.74

 pencil

1.71

Pen

1.70

 Pens

1.68

 fountain

1.66

Pen

1.60

Activations Density 0.175%

phrases indicating personal experience or preference in a positive context

various elements related to storytelling and narrative structure

No Comments

No Known Activations