INDEX

Explanations

terms related to different styles, such as fashion, cooking, music, and coding

oai_token-act-pair · gpt-3.5-turbo

references to different styles

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

orb

-0.76

ajor

-0.71

ertodd

-0.67

foreseen

-0.67

minster

-0.66

enary

-0.66

MS

-0.65

 Lauder

-0.65

riel

-0.64

eor

-0.64

POSITIVE LOGITS

ologies

0.95

heet

0.85

 sheets

0.80

 cues

0.78

 preference

0.76

faces

0.74

face

0.73

 sheet

0.73

uality

0.72

 flair

0.71

Activations Density 0.041%

terms related to different styles, such as fashion, cooking, music, and coding

references to different styles

No Comments

No Known Activations