INDEX

Explanations

words related to official titles or locations

oai_token-act-pair · gpt-3.5-turbo

words related to various types or categories of substances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

rir

-0.77

EEK

-0.75

MENTS

-0.74

CHA

-0.72

ENDED

-0.71

lehem

-0.70

herty

-0.69

houses

-0.69

ivity

-0.68

CLASS

-0.67

POSITIVE LOGITS

ysis

1.28

ity

1.15

ogue

1.00

 piety

0.89

arial

0.87

ized

0.84

ities

0.84

ibrary

0.83

acies

0.76

ogy

0.74

Activations Density 0.038%

words related to official titles or locations

words related to various types or categories of substances

No Comments

No Known Activations