INDEX

Explanations

alphanumeric codes or technical details like ISBN numbers and model numbers

oai_token-act-pair · gpt-3.5-turbo

references to measurements or numerical information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

Nun

-0.85

Neh

-0.82

 improvis

-0.75

 volunt

-0.74

 Ã©

-0.73

Nou

-0.73

nex

-0.73

Nek

-0.73

Ney

-0.73

 neoc

-0.72

POSITIVE LOGITS

CAST

0.75

ilings

0.75

0.74

ARGET

0.71

uca

0.69

asaki

0.67

AMP

0.67

BILL

0.66

]=

0.65

AGES

0.65

Activations Density 0.358%

alphanumeric codes or technical details like ISBN numbers and model numbers

references to measurements or numerical information

No Comments

No Known Activations